Skip to main content

Full text of "Laboratory manual of elementary colloid chemistry"

See other formats









By J. A. Crowther, Sc.D. Second Edition. With 29 Illustrations. 6». net. 


By E. B. Maited, Ph.D. With 12 Illustrations, js. net. 


By Emil Hatschek. Third Edition. With 17 Illustrations, 


By R. S. Willows, D.Sc, and E. Hatschck. With 21 Illustrations. 
Second Edition. 5s. net. 


By W. P. Dreapkr, F.I.C. 2S. 6d. net. 


By J. W. HiNCHLEY, A.R.S.M., Wh.Sc. With 70 Illustrations. 2S. 6d. net. 


By Dr, W. Ostwald. Translated by Dr. M. H. Fischfr, with notes added 
by E. Hatschek. Second English Edition. With 93 Illustrations. i8s. net. 




With 20 Illustrations 




Printed in Great Britain. 



Although most of the existing text-books of Colloid 
Chemistry necessarily give, in more or less detail, descrip- 
tions of experimental procedure and instructions for making 
many of the classical preparations, no laboratory manual 
or collection of practical exercises such as has been found 
indispensable in the teaching of other branches of chemistry 
has so far appeared. The lack of such a work is all the more 
likely to check the spread of a practical knowledge of the 
disciphne, as many of the methods and materials of colloid 
chemistry are pecuhar, and strange even to students well 
trained in inorganic and organic chemistry. 

The present work is an attempt to fill this gap and to 
supply accurate and very detailed directions for carrying 
out the fundamental operations, for making a number of 
representative preparations, and for examining them by the 
standard methods. These are based throughout on personal 
experience of the processes described and of the difficulties 
experienced in teaching them. The examples chosen are, 
generally speaking, the simplest ones and, where alterna- 
tives are possible, those involving the smallest expenditure 
in apparatus and material. The task of selection has not 
been easy, and the attempt to delimit the elementary region 
of the whole domain may seem premature or arbitrary : the 
guiding principle has been to provide for the wants of those 
students of numerous branches of science who are finding 
some training in colloid chemistry an indispensable part of 
their equipment, and are able to devote a limited time only 
to acquiring its technique. 

For the guidance of readers desirous of going beyond the 
limits of this manual a number of references to recent 
literature are given at the end of each section. The papers 
quoted are mostly records of experimental investigations 
which are either alternative to, or more advanced than, the 
examples given in the text. 

Since the book is the first of its kind, the author wiU be 
very grateful for hints from readers who may find any of 
the directions given in it lacking in clearness or capable of 
being simplified. 

February, 1920. 



Preface ......... v 

Chapter I. — General Remarks on Apparatus, Materials 

AND Procedure ...... 9 

Choice of vessels, methods of cleaning the same. 
Making up solutions and sols. Filtration of sols. Dis- 
tilled water. Redistilled water. Variabihty of 
materials and degree of accuracy. General hints. 

Chapter II. — Dialysis . . . . . .16 

Graham's dialyser. Parchment bags. Arrangements 
for continuous flow. Parchment tubes. Parchment 
thimbles. Collodion thimbles. Flat collodion mem- 
branes and " Star dialyser." Dialysers for sols show- 
ing osmotic pressure. 

Chapter III. — Suspensoid Sols ..... 29 
Gold sols : reduction by tannin ; reduction by for- 
maldehyde. Palladium sol. Silver sol : reduction by 
dextrine ; reduction by tannin ; reduction by hydro- 
gen. Various methods of making gold and silver sols. 
Sulphide sols : cadmium sulphide sol ; arsenic sul- 
phide sol. Miscellaneous sols : Prussian blue sol ; 
ferric hydroxide sol. 

Chapter IV. — Suspensions ...... 37 

Mastic suspension. Other resins, dragon's blood, 

Chapter V. — Organosols ...... 39 

Reduction of silver in wool-fat. Other metals in the 

Chapter VI. — Emulsoid Sols and Gels ... 41 
Silicic acid sol, preparation from water-glass. Salts 
promoting setting of sol. Determination of Si02. 
Effect of lyotropic salts. Gelatin. Commercial raw 
material. Swelling and dispersion. Filtration of sols. 
Hardness : determination of melting and setting 
points. Examination of strained gels in polarized 
light. Agar. Swelling and dispersion. Filtration. 
Agar gel. Effect of lyotropic series on setting and 
swelling. Purified gelatin. Sols containing a definite 
amount per volume. 

Chapter VII. — Egg Albumin Sol .... 57 

Preparation and filtration of sol from dried albumin. 
Heat coagulation. Coagulation by adsorption. Salt- 
ing out and the Hofmeister series of anions. Reversal 
in acid sols. Heavy metal precipitation. Purifica- 
tion of sol made from dried albumin and from fresh 
white of eggs. CrystalUzed albumin. Dialysis of 
albumin sols to prevent dilution. 



Chapter VIII. — Emulsions ...... 63 

Pure oil-water emulsions. Electrol5rte coagulation and 
clearing. Concentrated emulsions. Preparation with 
alkali. Separation and phase-ratio. Use of soap solu- 
tions and simple apparatus for same. Phases of equal 

Chapter IX. — Ultra-Filtration ..... 69 
Apparatus and membranes for Bechhold's method. 
Ostwald's ultra-filters : for use with vacuum ; spon- 
taneous. Thimbles as ultra-filters. 

Chapter X. — Optical Methods of Examination . 76 

Arrangements for observing Tyndall cone and state of 
polarization. Ultra-microscopic and dark-ground 
examination. Jentzsch ultra-condenser. Dark-ground 
condensers. Method of cleaning sUdes and cover 

Chapter XI. — Cataphoresis ..... 81 

Simple U-tube apparatus. Nemst and Coehn's U-tube. 
Determination of velocity in unit gradient. Micro- 
scopic observation and measurement of cataphoresis. 
Method of preparing slide. Determination of velocity. 

Chapter XII. — Electrolyte Precipitation of Suspen- 

soiD Sols ....... 89 

Typical limit concentrations. Method of determining 
same. Titration. Precipitation with constant sol- 
concentration. Standard electrolyte solutions. Pro- 
cedure. Standard electrolyte solutions for positive sols. 

Chapter XIII. — Mutual Precipitation of Suspensoid 

Sols ........ 94 

Precipitation in definite ratios. Experimental proce- 
dure. Determination of optimum ratio of oppositely 
charged sols. Electric charge on non-precipitated 

Chapter XIV. — Protection ..... 96 

Precipitation in presence of protecting agent. Gold 
numbers. Experimental procedure for determining the 
same. Specific nature of protection. 

Chapter XV. — Viscosity Measurements . . . 100 

Types of capillary viscometers. Ostwald viscometer. 
Determination of relative viscosities. Use and dimen- 
sions of instrument. Series of viscometers with increas- 
ing bore of capillary. Cleaning. Necessity of constant 
temperature. Regulator and troubles experienced with 
toluene type. Elimination of density. Ubbelohde vis- 
cometer. Manostatic arrangement for same. Method 
of using apparatus. Concentration-viscosity and tem- 
perature-viscosity curves. 

Chapter XVI. — Adsorption (Qualitative Experiments) i 10 
Adsorption of dyes. Adsorption of lead salts. In- 
fluence of solvent. Electric adsorption. Selective 



Chapter XVII. — Capillary Analysis . . . .113 
Description of method. Detection of methyl orange in 
vegetable dye solution. Detection of picric acid. 

Chapter XVIII. — Determination of an Adsorption 

Isotherm . . . . . . .116 

Choice of oxalic acid. Experimental procedure. Con- 
centration of permanganate solution. Titration. 
Example of actual titration. Plotting y-C and log y- 
log C diagrams. Discussion and determination of ex- 
ponent. Comparison of different solutes. Difficulties 
of analytical methods. Attainment of equilibrium. 
Choice of adsorbent. 

Chapter XIX. — The Liesegang Phenomenon . .124 

Original formula for silver chromate rings in gelatin. 
Calcium phosphate in gelatin. Lead iodide in agar. 
Lead chromate in agar. Molar concentration. Indirect 
formation of rings. Reactions in silicic acid gel. Pre- 
servation of specimens. 

Name Index 132 

Subject-matter Index 133 




Chapter I. 


The apparatus employed in the operations to be 
described is, with very few exceptions, that available 
in any chemical laboratory. Glass vessels used for 
preparative work should, if possible, be of resistance 
glass ; this applies even to test tubes used for such 
work as experiments on electrolyte coagulation. 
Test tubes which turn distilled water containing a 
little phenolphthalein pink in a very short time are 
by no means uncommon, and should not be used 
for any purpose. As regards the choice of larger 
vessels, it should be remembered that very thorough 
cleaning is necessary, and that in many cases undue 
exposure of solutions to air is undesirable, so that the 
choice will fall on tall cylindrical beakers, conical 
beakers with spout, or Erlenmeyer flasks. Flasks 
with narrow necks are, generally speaking, undesirable. 

Vessels should be cleaned immediately after use, 
in any event, and again before use in the case of 
sensitive preparations. The methods to be adopted 
in the former case naturally depend to a great degree 
on the previous contents of the vessel. Suspensoid 
sols are, of course, easily washed off, although in 


some cases — especially with positive sols — small 
quantities are adsorbed on the glass surface so 
tenaciously that washing with dilute hydrochloric or 
nitric acid may be required to remove the adsorbed 
film. Very thorough and repeated washing is 
necessary after emulsoid sols ; it must be continued 
until the last water shows no trace of froth. Traces 
of gelatin, albumin, etc., allowed to dry in glass 
vessels are very troublesome to remove, and may 
require the use of nitric acid or hot dichromate- 
sulphuric acid mixture. Irreversible gels, e.g., silicic 
acid, formaldehyde-gelatin, etc., should be made 
only in vessels from which the gel can be easily 
removed, such as cylindrical beakers, preferably with 
thick walls. Whatever the method of cleaning, the 
vessels should finally be rinsed thoroughly with dis- 
tilled water and drained. Drying with cloths is to 
be avoided ; drying with alcohol and ether is only 
necessary in the case of small apparatus used for 
quantitative work, e.g., viscometers. Apparatus 
cleaned as described and kept with the opening down- 
wards will generally require only rinsing with several 
lots of distilled water before use. 

Operations like making up salt solutions of known 
concentration for the preparation of sols, electrolyte 
coagulation, etc., require the usual apparatus and 
call for no special remarks. Sols of emulsoids con- 
taining a definite amount of dry material to a given 
volume of dispersion medium also offer no difficulty. 
Substances which disperse in the cold, like albumin 
or gum arable, stick to the walls of the vessel in the 
earlier stages of the process, but are easily detached 
when swelling has progressed sufficiently. The use 
of thick-walled vessels is, however, advisable, as 
thin beakers are easily broken in trying to detach 
fragments which stick obstinately. If sols contain- 
ing a definite weight of substance in a given volume 
of sol are required, the procedure is a little more 


difficult. For gelatin it is fuUy described under that 
heading ; with materials like albumin, gum, etc., it 
will be found advisable not to make the sol directly 
in the measuring flask, as stirring is impossible. The 
weighed quantity of dry material should be placed 
in a beaker and dispersed with, say, 50 or 60 per cent, 
of the total volume eventually required, and the sol 
so obtained poured into the measuring flask. The 
beaker is then carefully rinsed with small successive 
lots of dispersion medium, which are poured into the 
flask : the aggregate volume of these washings must 
fall short of the mark by a few cubic centimetres. 
The flask is finally filled to the mark with the disper- 
sion medium from a pipette and the contents well 
mixed. This method is not quite accurate, but the 
error is generally not as great as that due to the 
variable moisture content of the starting material. 

Filtration will be necessary chiefly in the case of 
organic emulsoid sols. It is generally a somewhat 
tedious process and, whenever possible, should be 
left overnight. The residues which have to be re- 
moved are generally not crystalline, and the use of 
vacuum does not accelerate the rate of filtration 
materially beyond the first few cubic centimetres. 
Carefully folded filters may be used where only the 
filtrate is required, as the complete removal of residue 
from such filters is not easy. Ribbed or corrugated 
glass funnels utihze the paper surface better than 
smooth ones, but are not so easily cleaned. All 
funnels should have the spouts cut off quite short, 
say not more than 2 cm. below the cone, as the 
usual long spouts are difficult to clean. Fairly hard 
filter papers are advisable in most cases. 

Small quantities of troublesome sols may be 
filtered through shredded asbestos with good results. 
This can be used in the ball tubes fisted in most 
catalogues of chemical glassware in the following 
manner. A disc of silver foil is cut which will pass 



easily through the upper part of the tube (Fig. i), 
and this is perforated with a strong needle, being 
supported on a cork plate for the purpose. The disc 
is then placed in the lower part of 
the ball and pure shredded asbestos 
packed into it up to its junction with 
the tube. The asbestos must be intro- 
duced in small quantities, moistened 
with the solution to be filtered, and 
rammed down hghtly with a stirring 
rod ; the exact degree of pressure 
required can be found only by experi- 
ence. The tube is then filled with 
the liquid to be filtered and the first 
few cubic centimetres of filtrate re- 
turned to it, if turbid. Concentrated 
sols like those of albumin or gum 
arable, if filtered in this manner, will, 
of course, generally be still opales- 
cent, but will be sufficiently clear in 
moderate thicknesses to allow the 
effect of coagulants to be seen dis- 
tinctly, and will be free from particles 
which would interfere with, say, vis- 
cosity measurements. 

A centrifuge capable of dealing 
with at least lOO c.c. at a tim^e will 
be found a very useful piece of appa- 
ratus, but is not indispensable. 

A microscope provided with a |" and 
a I" objective and with at least one 
high-power eyepiece, is required for the 
examination of sols by dark ground 
and ultra-condensers. The special features of these 
appliances are fully dealt with in the section devoted 
to them, but a general knowledge of the microscope 
must be presumed. 

As regards materials, the most important one is 

Fig. I. 


pure distilled water. If a sufficient amount of con- 
ductivity water is available, all difficulties are 
avoided. It is, however, not essential for the work 
described in the following pages, and the extreme 
precautions taken in the case of a few preparations 
prominently mentioned in the literature have pro- 
duced a somewhat exaggerated impression of the 
standard of purity required for less delicate work. 
Water distilled with ordinary care in reasonably 
designed apparatus will answer for aU but a few pur- 
poses. Trouble is much more likely to be caused 
through the ordinary storage vessels, since they are 
very rarely made of resistance glass. If trouble is 
experienced, the first thing to do is to use freshly 
distilled water only, and to coUect the small quan- 
tities which wiU be required in resistance glass flasks. 
Storage vessels coated inside with paraffin wax may 
be used ; remember that the lining cannot be 
removed again by warming, as the glass cracks 
before the wax melts. 

For the few preparations which require exception- 
ally pure water small quantities may be redistilled 
and condensed in a silver cooler. A thin-waUed 
silver tube about f " bore is not expensive, and can 
easily be fitted to a Liebig condenser in place of the 
glass cooler. There must, of course, be no parts on 
which water can condense, leading down the cooler : 
in other words, the cooler must be bent down to the 
distilling flask. This may be done by filling the tube 
completely with fine, dry sand, corking both ends 
and bending slowly over a cyUndrical object of 3" or 
4" radius. The water should be redistilled from, and 
collected in, resistance glass flasks. 

It is hardly necessary to add that throughout this 
book water means distilled water of the standard 
quality available ; where redistilled water is^essen- 
tial, or tap-water permissible, this is specially 


As regards the other materials, the ordinary 
chemicals call for no remarks. The solutions made 
from them for such investigations as electrolyte 
coagulation are molar and not normal, and this should 
be borne in mind. If there is any doubt about 
crystals containing the full amount of water of 
crystallization, or if the salts are anhydrous but 
hygroscopic, e.g., AICI3 or NH4CNS, the solutions 
must be standardized by the usual analytical 

Materials Lifce gelatin, agar or dried albumin are 
not definite chemical individuals, differ sHghtly or 
even considerably when obtained from different 
sources, and contain an amount of moisture which 
varies perceptibly. The best that can be hoped for 
is concordant results ; the first essential for this pur- 
pose is to use the same material throughout a given 
investigation and, therefore, to start with a sufficient 
stock to allow for all contingencies. 

These considerations have an obvious bearing on . 
the degree of accuracy to be aimed at in weighing and 
measuring. Centi- or miUi-molar solutions of elec- 
trolytes may be made up with the same care as 
solutions for volumetric analysis, althoi^h the 
operations in which they are eventually used have 
not a sharp end-point. On the other hand, it is 
unnecessary (and will, fortunately, also be found 
impossible with many atmospheric conditions) to 
weigh 10 gm. of gelatin to fractions of a milligramme, 
since the moisture content may easily vary by 0*5 per 
cent, of the total weight in a very short time. No 
general rules can be given which would be an adequate 
substitute for the exercise of common sense in this 

A few general hints on procedure — many of which 
may appear superfluous to some reader or other, but 
have not been found so by the author — may conclude 
this introduction. 


Read the whole chapter before beginning any of 
the work described in it ; although the operations 
are generally put in the order in which they succeed 
one another, it is well to have a complete idea of the 
work before starting. 

Many preparations change with age ; do not make 
more than you require for immediate use or than will 
keep safely. 

Label all preparations immediately in terms 
which, if not entirely correct technically, will remain 
intelligible to yourself. This is particularly impor- 
tant in the case of series, like solutions of different 
concentrations, Liesegang preparations, etc. 

Adapt your methods to the pecuUarities of your 
material. For instance, when told to dilute a 5 per 
cent. coUodion with an equal volume of acetic acid, 
do not put the highly viscous sol in the measuring 
vessel first and -pour the thin solvent on it, but proceed 
in the reverse order. 

When an experiment fails, repeat it with the 
alteration of one factor at a time. If, e.g., a gold sol 
turns out purple instead of red, try first a fresh 
beaker, than a fresh carbonate solution, and so on. 

Chapter II. 

The cheapest and most convenient membrane for 
dialysing any but small quantities — say 50 to 100 c.c. 
— is parchment paper. This is readily obtainable in 
sheets or cut in squares of various sizes. It varies a 
good deal in permeability, and only an actual trial 
can decide whether a particular sample is satisfactory 
for a given purpose. As the paper is fairly brittle in 
the dry state it should be kept flat or rolled, but never 

The classical method of employing the parchment 
membrane is that used by Thomas Graham, whose 
dialyser may be found in all catalogues of chemical 
apparatus. It consists simply of a glass cyhnder 
open at both ends, one of which is provided with a 
rim or groove, over which the membrane is tied. A 
circular piece about 2" larger in diameter than the 
cyhnder should be cut and thoroughly soaked in 
water, placed centrally over the rim, carefully turned 
down over it all round, and then tied with a thin 
string. Unless this is done with care, leakage may 
take place through some of the folds formed ; before 
a — possibly valuable — 'Solution is placed in the 
apparatus it should, therefore, be tested by filling 
it with water and ascertaining that it does not 
escape, (^he same precaution applies to all dialysers 
to be described in this chapter !) The dialyser is sus- 
pended or supported in a vessel filled with water, 
which is changed from time to time, or renewed 


Parchment bags are preferable for larger quanti- 
ties, as a larger surface is obtained in the same space. 
They may be made as follows : Cut a regular hexagon 
and soak it thoroughly in water. Then place it 
centrally on the bottom of an inverted beaker or jar, 
the diameter of which is about one-third of that of 
the inscribed circle of the hexagon. Gently pinch 
radial folds from the circumference of the beaker to 
the corners of the hexagon and mould them so that 
the paper midway between the corners touches the 
wall of the beaker, and then turn the folded portions 
over and smooth them into cylindrical shape. The 
whole procedure will be quite clear from Fig. 2, 
which shows the initial hexagon (dotted) and the 
final outHne of the edge in plan, as well as a per- 
spective view of the nearly completed bag on the 
beaker. The folds must not be sharp, as even wet 
parchment may be damaged by too drastic treat- 
ment. When the bag has been moulded as described, 
a string is loosely tied round it, or a fairly slack 
rubber band shpped over it within about 2" of the 
edge, and the bag is then drawn off the beaker. Its 
permanent shape is secured by threading a clean, 
thin string through the folds, as indicated by dotted 
line in the plan, which is gently drawn tight after 
every completed stitch so that the circumference at 
the open end is approximately the same as at the 
bottom. The bag is suspended in a jar of suitable 
size by two or three strings tied at equal distances to 
the string which secures the circumference. The jar 
is then slowly filled with water, while the liquid to be 
dialysed is poured into the bag at the same time and 
at about the same rate, so as to keep the external and 
internal level nearly the same ; in this way any 
strain on the mouth of the bag is avoided and it 
retains its shape. The water may be renewed from 
time to time, but it is preferable to use a continuous 
flow, as dialysis is greatly accelerated thereby. This 


Fig. 2. 


may be done by allowing water to flow into the outer 
vessel and removing it by means of a syphon, which 
must be of the type shown in Fig. 3, to avoid either 
the vessel or the syphon being emptied, if the water 
supply fails by any accident. The rate of supply 
must, of course, be so adjusted as not to exceed the 
rate of discharge from the syphon, since otherwise 
the water may flow over the top of the jar. It is 
hardly necessary to add that the same arrangement 
may be used with a Graham dialyser, and also, 
slightly modified, with many of the apphances yet to 
be described. Continuous flow can, of course, be 
used only when the hquid remaining in the dialyser 
is all that is wanted ; if it is, for any reason, necessary 
to examine the solution which has diffused through, 
dialysis must be performed with successive lots 
of water, which may be kept separate or be 

Parchment paper may also be obtained in the 
form of tubes — " sausage-skin dialysers " as they are 
usually termed in catalogues. They are sold flat, 
and in that condition are from 40 to 100 mm. wide, 
giving a diameter, when filled with Hquid, of 25 to 
70 mm. As they are easily damaged, any length 
selected for use should be carefully tested for leaks. 
It may be used in various ways : one of the simplest 
is to bend a (thoroughly soaked and tested) piece 
into U-shape and place it into a tall cyhnder, allow- 
ing the open ends to project an inch or two. The 
tube is then slowly filled with the Hquid to be dialysed, 
while the cyHnder is at the same time filled with 
water at about the same rate, so that no strain is 
placed on the tube. Another method is to close one 
end of the tube by folding it over two or three times, 
the first fold being about 5 mm. wide, and securing 
this end with a rubber cHp. The cHp is made by 
cutting a rectangular strip about 20 mm. wide and 
about 25 mm. longer than the width of the (flat) tube 



Fig. 3. 



from white rubber sheet about 8 to lo mm. thick, in 
which a central straight incision is made about 5 mm. 
longer than the width of the tube. This is then 
opened a httle by inserting two thin pieces of stick 
at the ends, shpped over the folded end, and then 
closed by withdrawing the sticks. When the use of 
metal is unobjectionable one of the wire chps used 
for attaching papers to one another may be em- 
ployed, or a similar chp bent from heavy silver or 
copper wire. The tube 
should be tested for leak- 
age after being closed. 

Finally, seamless thim- 
bles of parchment papers 
can be obtained, which, 
although somewhat ex- 
pensive, are rehable and 
extremely convenient, es- 
pecially for the examina- 
tion of small quantities 
of Uquid. A simple 
method of using them is 
shown in Fig. 4. The 
thimble, filled with the 
solution to be dialysed, is 

Fig. 4. 

placed in an Erlenmeyer flask of suitable size filled 
with the solvent. The parchment swells perceptibly 
in water, and the neck of the flask must, therefore, 
be a few millimetres larger in diameter than the 
(dry) thimble, to permit its easy withdrawal when 
it is saturated. 

A number of natural membranes, such as gold- 
beater's skin, fish -bladder, etc., have been used for 
dialysis. Since they vary in permeabihty or require 
careful purification, their use can hardly be recom- 
mended except as makeshifts, especially in view of 
the comparative ease with which membranes of con- 
siderable uniformity and covering a great range of 


permeability can be made by the methods to be now 

These are based on the use of collodion, i.e., sols 
of cellulose nitrate in a mixture of ether and alcohol. 
The raw material is obtainable commercially as " gun 
cotton " or " pyroxylin," and is generally sold damped 
with alcohol ; it should be dried before weighing. 
The usual concentration is 3 to 4 gm. of gun cotton 
to 100 c.c. of ether-alcohol mixture, the proportions 
of the latter varying between 14 parts of alcohol (90 
per cent.) to 86 of ether, and 25 parts of alcohol to 
75 parts of ether ; equal volumes of alcohol and 
ether have also been used, but this composition is 

The weighed quantity of gun cotton is placed in a 
wide-necked bottle, the requisite volume of alcohol 
poured on it, the bottle corked and allowed to stand 
for about fifteen minutes. The ether is then added and 
the mixture stirred occasionally, until the gun cotton 
has dissolved ; it should do so without leaving 
any residue. The sol should be almost clear and 
does not require filtering. One of the most con- 
venient ways of employing it is to make dialysing 
thimbles by coating the inside of test tubes of suitable 
size ; the beginner will find 20 mm. diameter X 
125 mm. long a convenient size, although with 
practice much larger thimbles can be made without 
difficulty. The test tubes must be quite smooth on 
the inside, thoroughly clean and dry. 

The selected test tube is filled with collodion, care 
being taken to pour it down the side so as not to form 
any air bubbles. The mouth of the test tube is then 
placed above that of the bottle and the collodion 
poured back slowly by slightly incUning the tube and 
rotating it constantly and slowly. The inclination 
of the tube is gradually increased as emptying pro- 
ceeds, but not more than is necessary to allow the 
collodion to flow out in a thin uniform stream. If it 


is raised too rapidly the bottom of the thimble is apt 
to be excessively thin. The tube is finally brought 
to a vertical position and the last remains of collodion, 
which should not then amount to more than a few 
drops, allowed to drip off, after which the layer left 
on the inside of the tube is allowed to dry for a short 
time. Although the degree of drying is the crucial 
point of the whole process, no definite rules can be 
given ; the collodion should not stick to the finger 
when touched lightly, and should just be visible as a 
faintly bluish coating when the tube is viewed 
against a dark background. When this stage is 
reached the tube is submerged in water, care being 
taken to allow all air to escape, and is left for at least 
15 minutes. The depth of water should be about 
2" more than the diameter of the tube, so that the 
subsequent operation can be carried out without its 
being uncovered. After the minimum time of 
immersion has elapsed, the collodion film is detached 
round the edge of the tube, a finger inserted so as to 
touch the collodion skin, and the latter very slowly 
pulled out, while the test tube is held with the left 
hand. It must be remembered that the rate at 
which the collodion skin can be pulled out is fixed by 
the rate at which water can flow through the space 
between it and the waU of the test tube, which is 
necessarily slow ; any attempt to hurry matters is 
fatal. If the bottom of the vessel containing the 
water is dark, the collodion membrane can be seen 
very distinctly, and the bottom end, which is the 
most likely portion to give trouble, watched. 

The finished thimbles can be kept under water for 
several weeks, undergoing only slight changes in 
permeabiUty. On the whole, however, it is advis- 
able to make and use them fresh. They can be 
mounted in a variety of ways ; a convenient method 
is to insert a short piece of glass tubing, the edge of 
which has been carefully rounded in the flame, and 


to fix the thimble to it with collodion, or by tying ; 
in the latter case a strip of gutta-percha tissue or 
oiled silk must be wound over the collodion to 
prevent it from being cut by the thread used for 

As has already been pointed out, the permeabiUty 
of thimbles made from ether-alcohol coUodion 
depends very largely on the extent of drying which 
they have undergone before immersion. The ether 
and alcohol still remaining in the film is replaced by 
water, and this fixes the permeabihty of the hydrogel 
of cellulose nitrate which ultimately constitutes the 
membrane. Although practice soon enables a care- 
ful worker to turn out fairly uniform thimbles, the 
whole difficulty can be avoided by the use of acetic 
acid collodion. This is made by dissolving 4 gm. of 
gun cotton in 100 c.c. of glacial acetic acid ; lower 
concentrations give fragile films, while higher ones 
produce unnecessarily dense membranes. The test 
tubes are coated and the excess emptied in exactly 
the same way as described ; the film is, however, not 
allowed to dry, but the coated tubes are immediately 
submerged in water. After about 30 minutes they 
may be withdrawn as explained, the operation 
being generally easier than with ether-alcohol collo- 
dion ; they are then left in water, which is occa- 
sionally changed, until the whole of the acetic acid 
has diffused out, and may be preserved under water. 

A very convenient way of making dialysing 
thimbles, which are less fragile and permit much 
greater variations in permeability than those just 
described, consists in impregnating the seamless 
thimbles of filter paper (Soxhlet thimbles) made in 
various sizes for fat extraction. One of these is held 
vertically over a small dish and filled to the top with 
collodion ; when the liquid has penetrated over the 
entire surface, it is inverted and drained with con- 
stant turning. If acetic acid coUodion is used, the 


thimble is then submerged in water immediately ; 
with ether-alcohol collodion it must, Hke the thimbles 
formed in test tubes, be allowed to dry for a few 
minutes before immersion. Since the mechanical 
strength is provided by the filter paper, collodions 
of low concentration may be employed, 2 per cent. 
in either acetic acid or ether- alcohol being sufficient 
for most purposes. The thimbles are strong enough 
to stand upright, and may be used like parchment 
paper thimbles. A convenient method of using any 
type of thimble is to stand or suspend it in a cylin- 
(Mcal vessel of sUghtly larger diameter, provided 
with an inlet at the bottom and an overflow outlet at 
a level i to 2 cm. below the top edge of the thimble. 
Water is continuously passed in at the bottom and 
overflows at the top, and dialysis proceeds with great 
rapidity with comparatively small quantities of water. 
Flat membranes of. (ether- alcohol) collodion are 
rather easier to make than thimbles, and can con- 
veniently be used for continuous dialysis in the " Star 
Dialyser " described by Zsigmondy. The apparatus 
(Fig. 5) consists of two parts, both of ebonite, a disc 
provided with a rim about 10 to 15 mm. deep, and a 
cylinder which fits loosely into the latter, open at both 
ends and 30 to 40 mm. deep. The disc has a central 
inlet and its upper face is provided with six or eight 
ribs, about 3 mm. deep, which stop a few millimetres 
short of both the central opening and of the rim. 
To prepare the membrane the ring is placed on a 
clean piece of plate glass and sufficient collodion 
poured into it to cover the glass to a depth of 2 or 
3 mm. The ring is hfted sUghtly, to allow the 
collodion to penetrate between it and the glass ; 
to strengthen the joint thus made, the outside of 
the ring is painted with collodion to a height of 
about 5 mm. from its lower edge. After the collo- 
dion has dried some minutes water is poured into 
the ring, which, together with the collodion mem- 



brane adhering to it, can be lifted off the glass 
after about lo or 15 minutes. The ring is then 




m//;j//////w//w///jM ' "-^ //////////////////mrnmA 


Fig. 5. 

placed into the disc, filled with liquid to be dialysed, 
and water is passed through the central inlet, which 



overflows round the edge of the rim. As it is diffi- 
cult to adjust the apparatus so exactly that overflow 
is uniform all round the rim, it is best 
to locaUze it by means of two or three 
strips of filter paper, placed between 
the open cyUnder and the rim and bent 
over the latter, so as to act as syphons. 
Special arrangements are necessary 
when the sol to be dialysed has an 
appreciable osmotic pressure, as is the 
case, e.g., with albumin sols. In this 
case water flows into the dialyser, 
diluting the sol and eventually causing 
it to overflow. The only way to prevent 
this is to counterbalance the osmotic 
pressure hydrostaticaUy ; in other 
words, to keep the level of the sol in 
the dialyser above the water level out- 
side from the beginning. For small 
quantities, such as come into question 
here, the simplest arrangement is that 
shown in Fig. 6. A dialysing thimble 
of either parchment or paper impreg- 
nated with collodion is fitted with a 
rubber stopper and tied tightly ; a 
strip of gutta-percha tissue about 
15 mm. wide is first wound round the 
end of the thimble and strong thread 
tied over this. Through the rubber 
stopper passes a funnel tube about 30 
cm. long, which must have a diameter 
of at least 8 mm., so that the sol can 
be poured down one side of it, allowing 
the air to escape and the thimble and 
tube to be filled to the top. The 
thimble is submerged in a beaker 
through which water flows con- 
tinuously. With this arrangement fig. 6. 


the liquid remains at its original concentration and 
the bulk of it is still contained in the dialysing 
membrane, as the volume of the funnel tube is 
comparatively small. 


Membranes made from collodion of unusually high 
concentration and capable of standing considerable 
pressures are described by A. T. Glenny and G. S, 
Walpole, Biochem. Journ., IX., 284 (1915) ; G. Wegelin, 
Koll.-Zeitschr., XVIII. , 225 (191 6), a new method of 
rapid dialysis and ultra-filtration. 

Chapter III, 


A. Metallic Sols. 

Gold Sols. — A I per cent, solution of gold chloride 
(more correctly, auro-chlorhydric acid, HAuCl^ . 
3H2O) serves as the starting material. The " gold 
chloride " of photography, NaAuCl4 . 2H2O, may be 
used instead for all the methods, with the exception 
of Zsigmondy's ; a i per cent, solution of this salt is 
obtainable in commerce. For the method first 
described, reduction by tannin, the gold chloride 
must be made exactly neutral to htmus by addition 
of sodium or potassium carbonate (N/5 solution). 

Reduction by tannin (Wo. Ostwald). Dissolve 
o-i gm. of purest tannin in 100 c.c. of water. If this 
solution is to be kept it should receive an addition 
of a few drops of chloroform, without which it goes 

Dilute I c.c. of the gold chloride solution with 
200 c.c. of water, stir and add i c.c. of the tannin 
solution, then warm over a Bunsen burner. Reduc- 
tion gradually proceeds and the Hquid becomes red. 
Continue heating and, when the hquid boils, add 
another cubic centimetre of gold chloride solution, 
followed by a cubic centimetre of tannin. The 
resulting sol should be perfectly clear in transmitted 
light and of deep ruby -red colour. The mixture 
must be well stirred after every addition. 

Sometimes the colour of the Hquid containing the 
first lot of gold chloride and tannin does not become 
red while warming, but purple, or even a cold violet. 


Do not be deterred, but continue to heat to boiling ; 
after the second addition of gold chloride and tannin 
at boiling point the colour very generally changes to 
red without even a tinge of purple. 

Reduction may also be carried out in the cold by 
using a larger proportion of tannin solution, say 
100 c.c. of water, i c.c. of gold chloride solution, and 
3 to 5 c.c. of tannin solution, added gradually. In 
this case the sol is more liable to have a purple or 
bluish tinge. 

Sols made by these methods are Hable to the 
growth of mould on keeping and gradually lose 
colour, the gold being deposited on the mycelium of 
the mould. This trouble may be prevented by 
adding a few drops of chloroform, or, in view of the 
great simplicity of the method, by preparing the sols 
when and as required. These sols are protected to 
a slight and uncertain extent by the tannin and its 
oxidation products, and are less suitable for coagula- 
tion experiments than the following. 

Reduction by formaldehyde (R. Zsigmondy) . Heat 
120 to 150 c.c. of redistilled water in a 300-c.c, 
beaker ; while it is warming add i c.c. of gold 
chloride solution (i per cent.), and then 2-5 to 3 c.c. 
of a N/5. solution of purest potassium carbonate. As 
soon as the solution boils stir vigorously ; add 
gradually, but fairly quickly, 2 to 3 c.c. of dilute 
formaldehyde solution (i c.c. of commercial 40 per 
cent, formalin to 100 c.c. of water) and extinguish 
the flame. Reduction is complete in about a 
minute, and the resulting sol should be perfectly 
clear in transmitted Ught and of pure ruby-red 
colour without purple tinge. 

The beaker used should be of resistance glass, and 
stirring rods of the ordinary soft glass must not be 
used ; a tube of resistance glass closed at one end 
should be used for stirring. The sol should also be 
kept in vessels of resistance glass. 


For some reason, which is still obscure, larger 
batches than about 150 c.c. cannot be made successfully. 
If larger quantities are required, they must be made 
in 150 c.c. lots as described ; since all the solutions 
can be made up in large quantities and keep indefi- 
nitely, there is no difficulty in preparing any volume 
of sol likely to be required. 

The sol can be dialysed against redistilled water, 
but will keep without this being done in vessels of 
resistance glass. It is very suitable as a standard 
preparation for experiments on electrolyte coagu- 
lation, protection, etc. 

Palladium Sol. — This can be prepared by exactly 
the same procedure as Zsigmondy's gold sol, using 
the following quantities : 150 c.c. of water, i c.c. of 
I per cent, palladium chloride solution, and 0-4 c.c. 
of N/io sodium carbonate solution, reduced by 3 to 
4 c.c. of dilute formaldehyde {i c.c. of commercial 
formahn to 100 c.c. of water). The sol should be 
brown and perfectly clear in transmitted light. 

Silver Sol. — Reduction by dextrine (Carey Lea's 
method). This is one of the best examples of a 
highly concentrated metaUic sol, as concentrations 
up to 5 per cent, of Ag can be obtained in favourable 
conditions. The following quantities should be tried 
first, but there is no difficulty in deahng with four or 
five times these amounts. 

Dissolve 4 gm. of commercial dextrine in 100 c.c. 
of water and then 4 gm. of purest caustic soda. 
Dissolve 3 gm. of silver nitrate in 20 c.c. of water 
and add to the dextrine-soda solution. A precipi- 
tate of silver oxide forms, which is gradually reduced 
by the dextrine, the colour changing to a reddish- 
brown. Allow 20 to 30 minutes for this, and then 
add 100 c.c. of 96 per cent, alcohol and stir. Allow 
the mixture to settle for another 15 to 20 minutes, 
and then pour off the turbid hquid from the sediment 
of silver as completely as possible. On pouring on 


water the silver generally disperses immediately ; 
should this not be the case, a Uttle shaking and 
stirring will be sufficient to induce dispersion. 

The silver amounts to i-8i gm., and in favourable 
conditions 35 to 40 c.c. of water will disperse the 
whole of it, so that the sol contains about 5 per cent, 
of disperse phase. It is, however, advisable to use 
a greater volume of water, say about 180 . .c. This 
sol is dark brown and opaque even in thin layers ; 
when diluted with about 50 times the amount of 
water it should be clear in transmitted hght, with a 
greenish-black surface colour in reflected light. The 
colour of the sol, and in fact the success of the whole 
method, depends a good deal on the quality of the 
dextrine, which can be determined only by experi- 
ment. Generally speaking, the ordinary yellow 
commercial brands work better than a highly purified 
product. The i per cent, sol may be kept unaltered 
for a long time. 

Reduction by tannin. Add to 100 c.c. of water 
I c.c. of I per cent, silver nitrate solution, and then 
a few drops of weak ammonia. Reduce with 3 to 
4 c.c. of 0*5 per cent, tannin solution. The sol should 
be brown and perfectly clear in transmitted hght, 
with a marked green surface colour in reflected Hght. 

Reduction by hydrogen (Kohlschuetter's method). 
This process is of interest as giving an electrolyte- 
free sol. Dissolve i gm. of silver nitrate in 20 c.c. 
of water and precipitate with a slight excess of 
caustic soda. Wash the precipitate of silver oxide 
by repeated decantation with hot water, and then 
suspend it in 200 c.c. of redistilled water, shake well, 
and then filter off any undissolved oxide. Pour the 
solution into a resistance glass flask kept at 
56° to 60° C. in a water bath (or thermostat), and 
pass a current of hydrogen through it b}'' means of 
a tube of resistance glass. Reduction is complete in 
20 to 25 minutes. 


Other methods. As gold chloride is reduced by 
most reducing agents, a very large number of 
methods of preparation of gold sols are possible ; 
references to some of these are given at the end of 
this chapter. Generally speaking, sols wiU result if 
gold chloride solutions containing about one part in 
10,000 are treated with small quantities of the 
following reducing agents, in solutions containing 
from one part in 4,000 to one in 500 : galUc acid, 
hydroquinone, pyrocatechin, white phosphorus in 
ether (Faraday's method, developed by Zsigmondy, 
q.v.), hydrazine hydrate (Gutbier), phenylhydrazine 
hydrochloride (Gutbier and Resenscheck) , etc. Silver 
may similarly be reduced from dilute silver nitrate 
by ferrous citrate (Carey Lea), in alkahne solution 
by hydrazine hydrate, all photographic developers, 

B. Sulphide Sols. 

Cadmium Sulphide Sol. — This is an instance of a 
sol produced by peptisation of a coarse precipitate. 

Dissolve 0-5 gm. of cadmium chloride in 20 c.c. of 
water and precipitate with moderately concentrated 
ammonium sulphide. The precipitate should be a 
deep yellow and should settle rapidly ; if it does not, 
the ammonium sulphide solution requires diluting. 
Wash the precipitate by decantation with two or 
three lots of water, 50 c.c. each, and suspend in 300 
to 400 c.c. of water. Pass a slow stream of hydrogen 
sulphide through the mixture and shake occasion- 
ally. The suspension first becomes milky, then 
yellow and moderately clear, and after 20 to 25 
minutes most of the precipitate will have been ''dis- 
persed. The sol may be filtered to remove any 
remains of precipitate, and boiled to drive off the 
excess of hydrogen sulphide, without coagulation 
occurring. The filtered sol is a pale golden yellow 

L.M, 3 


in transmitted light, with marked greenish opales- 
cence in reflected Hght. 

Arsenic Sulphide Sol. — This sol has been the sub- 
ject of many classical investigations, especially on 
electrolyte coagulation. To prepare it, dissolve 
2 gm. of arsenic trioxide in one htre of water ; keep 
the latter boiling until solution is complete. After 
cooUng the liquid pass a slow stream of hydrogen 
sulphide through it, with occasional stirring, until the 
colour does not deepen perceptibly. The sol is a 
pale orange colour in transmitted light, with a 
greenish-yellow opalescence in reflected light. Excess 
of HgS can be removed by passing hydrogen through 
the sol ; this must be done if the sol is to be used for 
coagulation experiments. 

C, Miscellaneous Preparations. 

Prussian Blue Sol. — Dissolve 0-4 gm. of crystal- 
lized potassium ferrocyanide in 20 c.c. of water and 
0*4 gm. of ferric chloride in 20 c.c. of water. Pour 
the first solution into the second, slowly and without 
stirring. Allow the mixture to stand for a few 
minutes and then pour it on a folded filter of hard 
paper. The filtrate should be quite clear and run 
fairly freely. When filtration is complete, wash the 
precipitate with four successive lots of 25 c.c. of 

The precipitate is then dissolved in 300 c.c. of 
solution containing 16 gm. of crystallized oxalic acid. 
The simplest way to do this is to pour the acid 
solution on the filter and allow it to percolate ; the 
precipitate will be found to have been completely 
dissolved when the whole volume of acid has passed 
through the filter. The solution of Prussian blue in 
oxahc acid is then dialysed in a parchment bag 
against repeated changes of distilled water, until the 


last batch of the latter gives no perceptible oxalate 
reaction. Owing to its deep colour and sensitiveness 
the sol is very suitable for cataphoresis and coagula- 
tion experiments. For the latter purpose the sol 
can be diluted with an equal volume of water, as 
even in that dilution it is deeply coloured in a thick- 
ness of I cm. The concentrated sol is quite stable, 
and there is, therefore, no reason for making a dilute 
sol directly, as this course entails waste of oxalic 

Ferric Hydroxide Sol* — Heat 500 c.c. of water in 
a tall beaker and, when it is boiling vigorously, add 
2 c.c. of a 30 per cent, solution of ferric chloride, 
gradually and with stirring. The Uquid turns a 
deep reddish-brown and remains perfectly clear. 

The sol contains HCl, corresponding to o-6 gm. of 
FeClg, i.e., about 0*4 gm., or approximately 22 miUi- 
moles per Utre. As this is a small fraction only of the 
HCl concentration required for coagulation, the sol 
may be used for precipitation experiments without 
being dialysed, as well as for cataphoresis in the 
U-tube (no particles are visible with dark-ground 
iUumination, so that the microscopic method is not 
appHcable). Most of the HCl can be removed by 
dialysis in the parchment bag, but only experience 
will tell how far dialysis may be continued without 
coagulation of the sol. Both the acid and the 
dialysed sol keep indefinitely. Only the latter is 
suitable for experiments on the mutual coagulation 
of oppositely charged sols. 


For aU inorganic suspensoid sols : Th, Svedberg, " Die 
Methoden zur Herstellung kolloider Loesungen anorgani- 

* Although this sol has some emulsoid properties, it is classed 
here with the suspensoids on account of its behaviour to elec- 
trolytes, etc. 



scher Stoffe," Theodor Steinkopff, Dresden, 1909. Sols 
with unusual electric charges : H. S. Long, Proc. Univ. 
of Durham Phil. Soc, V., Part 2 (1913), positive red gold 
sol; F. Powis, Journ. Chem. Soc, 107, 818 (1915), 
negative ferric hydroxide sol. Dye sol with charac- 
teristic suspensoid properties : Wo. Ostwald, Kolloid- 
chemische Studien am Kongorubin, Kail. Beihefte, X., 179 

Chapter IV. 

Mastic Suspension. — ^This preparation is one of the 
classical subjects of investigation. Dissolve o-i gm. 
of powdered gum mastic in lo c.c. of alcohol or 
acetone. Pour the solution slowly into 500 c.c. of 
water, stirring the latter vigorously. Filter the 
suspension through a fairly close filter paper to 
remove coarser particles. 

The preparation is almost opaque, with vivid pale 
blue opalescence, in reflected hght, and should be 
perfectly clear and a faint yellow in transmitted 
light. It shows an extremely bright blue Tyndall 

Coagulation experiments should be made with 
HCl and with salts of uni-, bi- and tri-valent cations, 
as the suspension behaves somewhat differently 
from suspensoid sols (see chapter on Electrolyte 
Coagulation). " Titration " will be found somewhat 
difficult, as there is no marked sudden change ; on 
standing, however, the disperse phase separates very 
clearly as a flocculent precipitate, though sedimenta- 
tion is naturally slow. In the U-tube a sharp boun- 
dary wiU be seen if observed in reflected light. 

A similar suspension, using exactly the same 
quantities, may be made from other resins. The 
beginner wiU find dragon's blood convenient, as the 
colour is a vivid red. 

Gamboge may be treated in the same way. A 
suspension which shows the Brownian movement, 
cataphoresis under the microscope, etc., can also be 


made by rubbing down a stick of the gum with a 
few cubic centimetres of water in a saucer (as is done 
with sticks of Chinese ink), diluting the resulting 
mixture with a large volume of water and filtering to 
remove coarser particles. 

Chapter V. 


The most convenient method of directly preparing 
organosols of the noble metals is that of C. Amberger, 
in which wool-fat (lanoline) is used as protective 

To prepare silver sol, dissolve 3-5 gm. of silver 
nitrate in 5 c.c. of water and add this solution in very 
small quantities at a time to 15 gm. of cold lanoline, 
incorporating it thoroughly with the latter by means 
of a pestle or a silver spatula. The success of the 
subsequent reduction depends on the completeness 
with which this is done. If any silver nitrate is left 
in the form of drops, the oxide and silver formed 
from them are of course not protected by the wool-fat 
and remain as a coarse insoluble residue when the 
latter is taken up in an organic solvent. Then add 
in the same way a solution of i gm. of sodium 
hydroxide in 5 c.c. of water. The mass turns first 
yellow and then brown, owing to the formation of 
silver oxide. On standing in the light the latter is 
reduced to silver ; the reduction is accelerated by 
gentle warming and by turning over the mixture 
from time to time, so as to expose the whole of it to 
the hght. After about six hours reduction is gene- 
rally complete, and the product is dissolved in 50 c.c. 
of chloroform. Fifty c.c. of petroleum ether and 
about 25 gm. of fresh granulated calcium chloride 
are then added — the latter to remove water, etc. — 
and the mixture allowed to stand for five to six hours. 
The solution, which should be a clear reddish-brown 


when diluted with abQut 20 volumes of solvent, is 
then poured off ; the solvent may be allowed to 
evaporate, leaving a mass of colloidal silver in wool- 
fat of the original salve-like consistency. This dis- 
solves easily in ether, petroleum ether, also in fatty 
oils and in paraffin. 

Organosols of gold, platinum and metals of the 
platinum group may be prepared in similar fashion, 
for which the original papers should be consulted. 


C. Amberger, Koll.-Zeitschr., XL, 97, 100 (1912), silver 
and gold ; XIIL, 310, 313 (1913) ; XVII., 47 (1915), 
platinum and metals of the platinum group. 

Chapter VI. 


A. Silicic Acid Sol and Gel. 

A convenient starting material is a solution of 
sodium silicate having a density i'i6, made by 
diluting the commercial water-glass syrup with 
freshly boiled distilled water. The ratio of syrup to 
water is best ascertained by preparing a small lot, 
sufficient for determining the density with a spindle. 
The solution may be prepared in large quantities, 
and should be kept in a bottle closed by a rubber 
stopper or a glass stopper well rubbed with vaseline. 

To prepare a sol, dilute 30 c.c. of concentrated 
hydrochloric acid (i'2 sp. gr.) with 100 c.c. of water, 
and pour 75 c.c. of the sodium silicate solution into 
the dilute acid. The mixture is dialysed in a parch- 
ment bag against repeated changes or against 
running water ; the beginner will find the former 
course more satisfactory. Experience will show how 
far it is possible to push dialysis without the sol 
setting to gel prematurely in the dialyser. 

The sol should be perfectly clear and colourless. 
It win keep for a length of time which can be ascer- 
tained only by experience ; as the removal of gel, 
formed accidentally, from flasks or bottles with nar- 
row necks is inconvenient, sol under examination 
should be kept in wide-mouthed bottles or taper 

Setting is greatly accelerated by COj, carbonates, 
phosphates, and free alkah. The effect can be 
demonstrated by bubbhng CO 2 gas through the sol 


until the bluish tinge, which indicates the beginning 
of gelation, appears ; or by adding small amounts of 
dilute solutions of carbonate, phosphate or ammonia 
to the sol, gradually and with constant stirring, 
which is discontinued as soon as the sol appears 
bluish. If the solutions are too concentrated, or are 
added too rapidly, local coagulation and flocculation 
may occur instead of complete gelation. 

The concentrations given above are fairly high 
and will be found useful if a stiff gel is required. If 
the sol alone is wanted and requires keeping for some 
time, the same quantities of hydrochloric acid and 
of sihcate solution should be used, but a larger 
volume of water. 

To determine the amount of SiOg in a given sol, 
evaporate 5 c.c. slowly in a weighed crucible to dry- 
ness and then ignite until the weight is constant. 
In the later stages of drying gelation may occur, and 
the steam bubbles formed in the gel burst violently 
and may scatter some of the material, unless drying 
proceeds very slowly. 

The effect of lyotropic additions is the same as in 
the case of other emulsoid sols. This can be shown 
qualitatively by placing 10 c.c. of freshly dialysed 
sol in each of three test tubes, keeping one as blank 
and saturating the others respectively with sodium 
sulphate and with ammonium thiocyanate. The 
sol containing Na2S04 will set before, and that con- 
taining NH4CNS after, the blank sample ; the latter 
very generally does not set at all. 

All vessels, measures, etc., used for sodium silicate 
or silicic acid sol should be washed immediately and 

B. Gelatin and Agar Sols and Gels. 

Gelatin occurs in commerce as " leaf " gelatin in 
sheets about 9" to 10" long by 4" to 5" wide, showing 


the diamond-shaped marks of the wire netting on 
which the leaf has been dried ; as powder, and as foil 
of uniform thickness — about 0-15 mm. — without any 
marks. The most suitable brands for practically all 
the work to be described are Coignet's " Photo- 
graphic " and " First Quality," and Nelson's " Crystal 
Leaf." Since different brands differ appreciably 
in their physical constants and in their ash content, 
it is essential to start any given investigation with 
an amply sufficient stock of the brand selected. If 
great constancy is aimed at, it is desirable to take 
leaves at random throughout a one-pound package, 
or to shear through the entire package and use the 
strips so obtained rather than the necessary number 
of adjacent leaves. 

For many purposes, i.e., in all cases in which only 
reproducible and not quantitative results are aimed 
at, sols and gels containing a definite amount to a 
given volume of water, e.g., 10 gm. of gelatin to 
100 c.c. of water, are quite suitable and are easier to 
prepare than sols containing a specified amount in a 
definite volume of sol. The leaf is broken into pieces 
preferably not larger than |" square, placed in a 
beaker, and the requisite amount of water poured on, 
care being taken that the whole of the leaf is covered ; 
air bubbles should be removed by shaking or stirring. 
The gelatin is then allowed to swell, either to com- 
plete saturation, or for any arbitrarily fixed period, 
which, however, should not be less than two or three 
hours. Complete swelling may take 24 hours or 
even more ; as gelatin imbibes something like ten 
times its weight of water, there wiU be no loose or 
unimbibed water, if the amount originally put on 
was less than ten times the weight of gelatin leaf. 
The thickened edges of the leaf take considerably 
longer to swell than the rest, and care should be 
taken that the time allowed is sufiicient to soften 
them completely. 


The next operation is the dispersing of the gelatin, 
which should be carried out on the water bath. A 
temperature of 35° to 45° C. is sufficient, but higher 
temperatures may be used to accelerate the process 
and for other reasons. Thus, if the sol is to be 
filtered (see below), it will be advisable to heat up to 
80° or 90°, as otherwise the viscosity is high and the 
rate of filtration excessively low. It is necessary to 
bear in mind that the properties of a gelatin gel or sol 
are not merely functions of the concentration and tem- 
perature, hut depend on its whole previous history, viz., 
the period allowed for swelling, the temperature at which 
the sol was formed and the length of time during which 
it was exposed to this temperature. To ehminate 
differences in the history, the practice is sometimes 
adopted of heating the sols for a definite time, say 
five minutes, to 100°, coohng at a definite rate to the 
temperature at which the sol is to be used [e.g., for 
viscosity measurements) and keeping the sol at the 
lower temperature, likewise for a definite time, 
before use. While this treatment goes a considerable 
way towards obhterating the " thermal history," it 
is yet safer to adopt a rigidly uniform procedure in 
any particular investigation. 

The sol in many cases does not require filtration 
and is ready for use when the gelatin is completely 
dispersed. If the gel is wanted especially for the 
study of its elastic or optical properties, it must not 
be used for at least four hours after setting is appa- 
rently complete, as the modulus and the accidental 
birefringence do not attain their final values before 
that time. Bodies of gel of definite shape can be 
made by pouring the sol into suitable moulds ; thus 
cyhnders can be made by using glass or metal tubes 
closed at one end, from which the gel cylinder is 
removed by dipping them into boiling water and 
allowing the gel to drop into an ample depth of cold 
water. Other shapes, e.g., prismatic ones, can be 


cast in moulds made from heavy tin or lead foil, or 
wooden moulds lined with ordinary tin foil, which is 
rubbed with vaseline, any excess being removed by 
wiping with cotton- wool. In all cases the gel should 
be left in the mould for several hours after setting, 
as mentioned above. 

In many cases, and always when a salt capable of 
forming a precipitate with calcium salts, chlorides, 
sulphites or sulphates has to be added, the sol will 
require filtering. The most suitable paper is Chardin's, 
either the original brand or an imitation made in 
England. It can be obtained in sheets or as folded 
filters, which, however, are too large for the smaU 
batches usually required. Folded filters should be 
made, great care being taken with the point ; as the 
paper is rather thick it is not advisable to try to make 
more than twelve folds. A hot-water funnel is 
used ; those usually obtainable have the defect that 
the spout of the glass funnel is much too long, so that 
cooling and even setting may take place in the por- 
tion which passes through the stopper of the water- 
jacket. To obviate this, a rather thin stopper, not 
more than |", should be used, and the spout of the 
glass funnel cut off so as just to project through the 
stopper. The temperature of the water bath should 
not be higher than is necessary to secure a reasonable 
rate of filtration ; this varies considerably with 
different brands of gelatin and, when solutes are 
present, with the nature of the latter. 

Gelatins are classified as " hard " and " soft," the 
former type being desirable for most investigations. 
The term " hardness " denotes a complex of quahties, 
among which are high " melting " and " setting " 
temperature and high elastic modulus. The melting 
and setting points are, of course, not strictly defined, 
and|can be determined and compared only by con- 
ventional methods. An apparatus suitable for this 
purpose is illustrated in Fig, 7. A test tube A is 



Fig. 7. 

suspended in the centre of a 
300014000.0. beaker B, which 
serves as a water bath, by 
means of the guide C, through 
which it must sHde freely. A 
tube I" diameter x 6" long is 
suitable ; it is weighted with 
15 to 20 gm. of mercury. It 
is essential that the tube 
should be perpendicular when 
it is resting on C ; if the rim 
is not sufficiently regular to 
ensure this, a square collar, 
say of rubber, should be used 
and permanently attached to 
the tube. A glass rod D, 
about f" diameter for a f" 
tube, is suspended exactly in 
the axis of the test tube. (If 
the apparatus is to be used 
frequently, it is advisable to 
mount it permanently to 
ensure correct alignment.) 

To determine the melting 
point the test tube is filled 
with a definite quantity of 
the gelatin sol under exami- 
nation, the beaker filled with 
water at a definite tempera- 
ture, say 15° C, and the sol 
allowed to set for a definite 
time. The rod, with the test 
tube hanging to it, is now 
raised a definite height (which 
stage is shown in the illus- 
tration), and the tempera- 
ture of the bath slowly raised, 
with constant stirring, until 


the test tube slides off the gel cylinder surround- 
ing the rod and comes to rest on C. The tempera- 
ture at this moment is noted as the " melting 
point." If the " setting point " is also to be deter- 
mined, the rod is lowered to its original position, the 
flame extinguished, and the bath allowed to cool. 
The rod is raised very slightly from time to time, 
until it just lifts the test tube with it, the tempera- 
ture at this point being noted as the " setting point." 
It must be remembered that there is considerable 
hysteresis and that the setting point of harder brands 
may be as much as 7° or 8° C. lower than the melting 
point of about 10 per cent. gels. 

A more delicate method of determining, with very 
simple means, the setting point is based on the well- 
known fact that the exposed surface of a gelatin gel 
which has been allowed to set quietly is 7Wf smooth 
like that of a liquid, hut shows a network of wrinkles. 
The formation of these wrinkles is not due to drying, 
but occurs actually during the last stage of setting. 
The alteration in the appearance of the surface is 
very striking if it is observed under an acute angle in 
reflected light, and it may be used for determining 
the setting point in the following manner : A small 
porcelain crucible is filled with about 10 c.c. of sol 
and the bulb of the thermometer completely immersed 
in the latter. The reflection of the window in the 
surface is then observed, attention being fixed on 
some dark object in the light field, such as the window- 
frame or the like. The reflection of such an object 
is, of course, distorted by the menisci formed by the 
sol at the wall of the crucible and the stem of the 
thermometer, but is a smooth and unbroken curve. 
As soon as wrinkling commences, the image is broken 
up into fringes (see Fig. 8, a and h) ; the fall of 
temperature between the time when this altera- 
tion in appearance becomes barely perceptible and 
when it is quite unmistakable rarely amounts to 



more than o-i°, which is a more than sufficient 

As regards other physical properties of the sol, the 
one most likely to require investigation is its vis- 
cosity at different concentrations and temperatures. 
The methods to be employed are described fully in 
the chapter deahng with viscosity measurements. 
Since viscosity is particularly sensitive to variations 



in the " thermal history," uniformity of procedure 
in the preparation of sols for this purpose must once 
more be insisted on as being of fundamental import- 

As regards the gels, reactions in gels are treated 
in a separate chapter. The quantitative study of the 
modulus of elasticity or the accidental birefringence 
produced by strain is beyond the limits of this book. 
It is, however, easy to demonstrate the latter by very 
simple apparatus, if a Nicoll and selenite plate are 



available, and the study, particularly of the strains 
set up by drying, is instructive. The apparatus is 
simply an open box (Fig. 9) about 16" high X 8" wide. 
Two glass plates — photographic plates from which 
the film has been removed 
are suitable — rest on the 
bottom, inchned under an 
angle of about 53° with the 
latter. A strip of either 
ground-glass or glass coated 
on the lower side with the 
" matt varnish " used in 
photography rests on two 
ledges, about 9" or 10" from 
the top of the box. An 
opening in the centre of the 
top takes the mount of the 
Nicoll prism, which can be 
rotated. If hght from a lamp 
placed as shown is reflected 
from the double glass plate, 
a sufficient fraction of it is 
polarized to show very shght 
strains in gelatin gels con- 
taining 10 per cent, and 

The strains set up during 
drying and their progressive ^ -' 
changes can easily be traced '- " 
and are instructive. A body 
of gelatin gel, unless it is a 
simple surface of revolution 
approximating fairly closely 
to a sphere, does not remain similar to itself during 
drying, and if the surfaces meet in edges very con- 
siderable distortion occurs. Thus, a right cyhnder 
with flat ends has two circular edges, and drying is at 
first much more rapid along these than it is on the 

Fig. 9. 



curved or flat surfaces. The edges, therefore, con- 
tract and the cylinder becomes a barrel with convex 
ends. The edges have now become so dry and rigid 
that very little further drying talces place in them, 
while the rest of the surface is rapidly shrinking, and 
the final shape is a single-shell hyperboloid with con- 
cave ends. Similarly, when a cube is allowed to 
dry, the edges contract first and the faces become 
convex, while the final surface has concave faces with, 
of course, concave edges. The distribution of strain, 
and the change from compression to tension, can 
easily be observed and analyzed. 

Gelatin being highly liable to putrefactive changes, 
neither sols nor gels can be kept for long without 
sterile precautions, which are beyond the scope of 
this work. Hardening agents like formaldehyde alter 
the physical properties of gelatin so much that they 
are suitable only for preserving finished specimens. 
Directions will be found in the chapter on the 
Liesegang phenomenon. 

Agar occurs in commerce as strips having a fibrous 
texture, as a fine powder, and as bars of square cross- 
section. The first-named is the cheapest form ; 
powdered agar has the advantage that the time 
necessary for swelling is considerably reduced. If 
strip is used it is torn into small pieces, which are 
allowed to swell in the requisite volume of water for 
about 24 hours. A small addition — one part in 
500 — of acetic acid is usual and promotes imbibition, 
but is not essential. The mixture is then boiled 
slowly, until the shreds have entirely disappeared. 
The sols are always turbid and show even macro- 
scopic fibres and fragments, so that they must at 
least be strained through fine muslin or through 
glass or cotton-wool plugs. If clearer sols and gels 
are required they must be filtered through Chardin 
paper in the manner described for gelatin ; the water . 
in the jacket must be boiling. With sols containing , 



I per cent, and over filtration is tediously slow, and 
the filtrate sets long before the filtration of even small 
batches is complete. If a large drying oven or 
sterilizer kept at 100° C. is available, the most con- 
venient course is to place the whole apparatus, i.e., 
filter funnel and beaker or flask for the filtrate, into 
it, when the whole can be left to itself without further 
attention. The setting temperature of agar sols is 
between 35° and 40°, while the melting point of gels 
hes between 90° and 100°, so that gels have to be 
heated on a water bath at boiling point to obtain a 
sol. Agar gels are not Uable to putrefaction and are, 
therefore, preferable to gelatin for long-continued 
experiments — e.g., on diffusion — in which the specific 
properties of the gel are of no consequence. They 
are, however, a good medium for the growth of various 
moulds and occasionally of BacUlus prodigiosus, 
which latter forms red patches. Both occur chiefly 
on the surface, and the rest of the gel may generally 
be used after the affected patches have been cut 

Agar gel, unlike gelatin, does not adhere to glass, 
and specimens may be removed from moulds with- 
out the heating necessary in the case of the latter. 
Thus cylinders may be cast in tubes stoppered at the 
bottom, and will drop out when the stopper is 
removed. A certain amount of hquid, which also 
contains agar, exudes from agar gels on standing, 
partly on the surface and partly between the gel and 
the containing vessel. This is a normal phenomenon 
and does not indicate faulty procedure in the pre- 

The Lyotropic Series. — It is desirable to demon- 
strate the general nature of the series by showing its 
effect on two sols as chemically different as gelatin 
and agar. Sulphates, chlorides and thiocyanates 
may be chosen as representative specimens, suffi- 
cient of each being placed into a loo-c.c. beaker to 



produce a concentration of N/2 in 50 c.c. of sol (allow 
for water of crystallization!). Each of the beakers 
so prepared now receives 50 c.c. of sol, 10 per cent, 
gelatin sol and i per cent, agar sol being suitable, 
and the same quantity of pure sol is placed into a 
fourth beaker for comparison. All four beakers are 
placed in the water bath until they have attained 
the same temperature and are then taken out and 
allowed to cool. The water bath should be at 35° 
to 40° C. for the gelatin sol and at boiling point for the 
agar sol. The order in which the gels set will be the 
same for gelatin and agar, and the intervals between 
the four specimens will be considerable ; the sol 
containing thiocyanate remains Uquid at room tem- 

The effect of the lyotropic series, and also that of 
dilute acid and alkaU, on the swelling of gelatin 
may also be demonstrated very simply in the follow- 
ing manner. Squares having a side of, say, 15 mm. 
are cut from the gelatin foil mentioned above ; if 
this is not obtainable leaf may be used, but the 
diamond markings, which are highly strained, should 
be avoided. The squares are placed in watch-glasses 
or Petri dishes containing a few cubic centimetres of 
the following solutions : N/50 HCl, N/50 NaOH, 
N/i Na2S04, N/2 NaCl, N/2 NH^CNS and water. 
The squares should be held in a small forceps and 
immersed quickly and completely, without allowing 
air bubbles to adhere to them. The difference in 
swelling will be quite noticeable after one hour (when 
the gelatin placed in the thiocyanate solution is 
probably completely dispersed), although complete 
equihbrium is not attained for many hours. It must 
be remembered that the foil swells in all directions 
and that the increase in volume is, therefore, pro- 
portional to the cube of the side. The squares are 
best examined by holding the watch-glass 2" or 3" 
above a black background, when they appear 


turbid, or of course by carefully pouring off the 

Purified Gelatin. — The work described so far can 
be carried out with the raw material obtainable com- 
mercially. This always contains electrolytes, which 
it may be necessary to remove as far as possible, 
although it should be noted that the physical pro- 
perties of the gelatin are sensibly affected by the 
prolonged washing which is required. A known 
quantity of the leaf is placed in a weiglied tall beaker, 
capable of holding a volume of water equal to at 
least 15 times the weight of gelatin. Running water 
is then passed in near the bottom of the vessel and 
allowed to overflow for 48 hours. If an adequate 
supply of distilled water is available it may be used ; 
faihng this, tap-water may be employed, in which 
case washing must be completed with several changes 
of distilled water. To prevent the formation of 
mould a few fragments of camphor or thymol, 
wrapped in mushn, are added between the leaves, 
so as not to escape with the water. 

Since gelatin imbibes something like 10 times its 
weight of water, sols of greater concentration cannot 
be made directly from the washed gelatin by warming 
to dispersion. To obtain them it is necessary either 
to dry the wet mass over H2SO4 or CaCL,, or to 
evaporate the dilute sol to the required concentra- 
tion at fairly low temperature, the concentration in 
either case being determined by weighing. The 
second procedure affects the sol perceptibly, especially 
if prolonged. 

Sols containing a Definite Amount per Volume. — 
We have so far dealt only with sols and gels contain- 
ing a definite percentage of gelatin to a given amount 
of water ; in other words, with sols containing a 
definite amount of gelatin in a given weight of sol. 
The preparation of sols containing a definite amount 
of substance in a given volume is comphcated chiefly 


by the fact that sols are not Uquid at the tempera- 
tures for which the usual measuring vessels are 
graduated, and the first point to decide is whether 
the sol is required to have a definite concentration 
at some particular temperature, say 35° for vis- 
cosity measurements, or at some arbitrarily chosen 
lower temperature, at which it may be transformed 
into gel. In the former case the flask to be used 
should be filled with water at the standard tempera- 
ture and then placed in a water bath and warmed to 
the temperature selected ; a fresh mark should be 
placed at the level reached by the water. The volume 
to this mark is calculated from the ratio of the 
specific volumes of water at the two temperatures 
selected ; for 15° and 35° C. respectively these are, for 
instance, 1-00085 a-^d 1-00586, so that the volume 
of, say, 500 c.c. measured at 15° wiU be 

500 X 1-00586 = 502-5 c.c. at 35°, 
which figure is noted and the gelatin content cal- 
culated on it. Since it is inconvenient to soak and 
disperse the leaf in a long-necked flask, this should 
be done in a beaker with about 75 or 80 per cent, of 
the total volume of water required, the more con- 
centrated sol thus obtained poured into the flask, 
which is placed in a thermostat at the required tem- 
perature, and the beaker washed out with successive 
small portions of warm water, which are transferred 
to the flask until the mark is reached. The contents 
of the flask, of course, require thorough mixing 
before use. 

Commercial leaf contains a considerable amount of 
moisture, rarely less than 10 per cent., which must 
be taken into account. It can be removed almost 
entirely by drying at 100° to constant weight (note 
that gelatin takes up moisture from the air even 
during the time required for weighing), which treat- 
ment, however, affects the properties of the material 


very considerably. It should, therefore, be applied 
only to a small sample, the rest of the material being 
kept in an airtight receptacle from the time at which 
the sample has been taken, so that its moisture con- 
tent remains constant. When making up sols to a 
given concentration per volume, this should, of 
course, be calculated on the weight of dry gelatin. 

It should be remembered that sols made by 
diluting a more concentrated sol differ slightly in 
their physical properties from sols produced by dis- 
persing the gelatin at once in the total water required. 
To obtain comparable results it is again necessary 
to observe the rigid uniformity of procedure which 
has already been insisted on, i.e., to use the same 
percentage for diluting, to have the added water at 
the same temperature as the sol, etc. 

If powdered gelatin (the usual brands of which are, 
however, markedly less " hard " than the best 
brands of leaf) can be used, the procedure is simpler, 
since there is no difficulty in carrying out soaking 
and dispersion in the measuring flask itself. The 
latter should be about half filled with water by means 
of a long (thistle) funnel, so that the neck remains 
quite dry. The powdered gelatin is then poured in, 
in small portions and in a thin stream, through a wide- 
necked funnel, and the flask shaken frequently to 
cause the powder to sink without the formation of 
lumps. On no account must the powder be placed 
in the dry flask first. When the whole of the powder 
is submerged further water is added to within about 
5 CO. of the mark^giving the volume at 15° C. (this 
refers to a 500-c.c. flask ; for other sizes the margin 
should be in proportion), and the necessary time 
allowed for swelling. The flask is then placed in the 
water bath, and the volume is made up to the mark 
giving the volume at the working temperature which 
has been fixed upon. Careful mixing before use is 
also required in this case, since the gelatin does not 


diffuse perceptibly during dispersion, so that a con- 
centrated layer of it rests on the bottom. 


A very complete Study of "Vapour Pressure, etc., of 
Gelatin-Water Systems," by K. Gericke, Koll.-Zeitschr., 
XVIL, 78 (1915) ; "Influence of Neutral Salts on Vis- 
cosity and Swelling," J. Loeb, Journ. of Biol. Chem., 34, 
77, 345 (1918). 

Chapter VII. 

The only commercial raw material is dried egg 
albumin, and the beginner should carry out the 
experiments described below with a sol made from 
it ; although its use is open to objections, the results 
obtainable correspond sufficiently closely to those 
recorded in the literature for " natural " albumin. 

Crush 15 gm. of dried egg albumin coarsely and 
introduce in small portions, with stirring, into 
100 c.c. of water. The albumin at first adheres to 
the walls of the vessel with great tenacity, but is 
easily detached as imbibition proceeds. The sol 
should be stirred from time to time and lumps 
broken up until dispersion is complete. The sol is 
turbid, with a varying small fraction of insoluble 
matter, which does not settle even on prolonged 
standing. It must be filtered, preferably overnight, 
through asbestos in the manner described on p. 12 ; 
filtration through paper is extremely tedious and 
involves considerable loss. The filtrate from asbes- 
tos is a yellowish Liquid, opalescent, but quite suffi- 
ciently clear in moderate thickness, say in test tubes 
15 or 18 mm. diameter, to allow even incipient pre- 
cipitation to be noticed easily. Albumin sols are 
very liable to undergo decomposition and should be 
used quite fresh ; in warm weather a trace of thymol 
may be added to the mixture before filtration without 
affecting the properties of the sol. 

Heat Coagulation. — Place a test tube containing 
about 10 c.c. of sol in a small water bath and heat 
slowly, stirring constantly with a thermometer. 
Note the temperature at which the sol begins to turn 


white and opaque. Remove the test tube from the 
bath, add water and break up the coagulum, to show 
that it does not disperse again, i.e., that the heat 
coagulation is irreversible. 

Irreversible Change by Adsorption. — Albumin is 
readily adsorbed at the interface between sol and 
another liquid, and becomes insoluble in the process. 
This is easily demonstrated by placing in a test tube 
10 c.c. of sol, adding i c.c. of some organic Uquid 
heavier than the sol, e.g., chloroform or carbon 
tetrachloride, and shaking vigorously, so that added 
liquid is broken up into small drops. These sink to 
the bottom and remain perfectly separate, which 
shows the formation of a film preventing their 
coalescence and evidently insoluble in the sol, as no 
change takes place even on standing for some time. 
To demonstrate that the adsorbed film is also 
insoluble in water, pour off the sol and replace by 
water ; no coalescence occurs even then. 

Salting-out and the Hofmeister Series of Anions. — 
It will be sufficient to try a few of the more charac- 
teristic salts of the series by adding the dry salts to 
the same volume of sol, so as to keep the albumin 
concentration approximately constant. The molar 
concentrations necessary to produce immediate tur- 
bidity in " natural " albumin sols [i.e., sols contain- 
ing, like ours, the other constituents of egg-white) 
are given below, as well as the amounts of the most 
readily obtainable salts required to produce these 
concentrations in lo c.c. volume : — 

M. per litre. Gm. in lo cc. of solution. 

Na citrate . 0-56 1-647 crystallized neutral sodium 


Na2S04 . 0'8o 2*577 crystaUized sodium sul- 

NaCHg . CO 2 1*69 2*298 crystalUzed sodium acetate. 

NH4CNS . — Ammonium thiocyanate to satu- 


Place the coarsely powdered salts in flasks with a 
mark at 10 c.c, or, if these are not available, in test 
tubes 15 or 18 mm. diameter which have been pro- 
vided with a mark at that volume. Place the flasks 
or test tubes in a water bath kept at about 35° and 
dissolve the salts gradually by slowly reversing the 
tubes at intervals ; they must not be shaken, as the 
formation of froth is to be avoided. Watch the 
appearance as solution proceeds and note that a 
marked turbidity appears only when the whole of 
the salt has gone into solution, except with thio- 
cyanate, which does not salt out even in saturated 

Dilute the turbid sols with an equal volume of 
water, and note that they become clear, i.e., the 
salting out is reversible. 

Reversal of the Hofmeister Series in Acid Sols. — 
Acidify 20 c.c. of sol by adding i c.c. of normal 
hydrochloric acid. Place in one test tube the same 
quantity of sodium sulphate (2-577 g^-) ^-s used 
above, and in another 3 gm. of ammonium thio- 
cyanate, fill to the 10 c.c. mark with the acid sol, and 
dissolve the salts gradually. No precipitate is formed 
in the sol containing the sulphate, while the sol con- 
taining thiocyanate becomes turbid and eventually 
clots completely. 

Precipitation by Salts of the Heavy Metals. — Pre- 
pare some 2N solution of copper sulphate (say 
12-480 gm. of the crystallized salt in 50 c.c. of 
solution). Add from a burette a few drops of this 
solution to 10 c.c. of sol ; a heavy greenish coagulum 
forms immediately. Continue to add copper sul- 
phate solution, with occasional stirring ; the pre- 
cipitate re-dissolves and has disappeared when 10 c.c. 
of solution has been added, i.e., when the mixture is 
normal in respect of copper sulphate. Now add 
sufficient powdered copper sulphate to saturate the 
20 c.c. (about 5-6 gm.) and dissolve ; a second 


precipitation begins when the solution has become 

To show that coagulation by salts of heavy metals 
is irreversible, repeat the experiment as far as adding 
a few drops of copper sulphate solution to lo c.c. of 
sol, then dilute with water, and note that the coagu- 
lum does not dissolve. 

The student desirous of working with pure albumin 
will do well to practise the usual methods of purifica- 
tion in the first instance with dried albumin. The 
methods are based on the fact that the constituents 
of white of egg other than albumin, viz., globulin, 
ovomucoid, etc., are salted out by lower concen- 
trations of ammonium sulphate than is albumin. 
Disperse 15 gm. of dried egg albumin in 100 c.c. of 
water, as described, but do not filter the sol. Add 
in small portions sufficient finely powdered pure 
ammonium sulphate to produce a half-saturated 
solution ; 38 to 39 gm. is required. Each addition 
should be made only after the previous one has dis- 
solved. A white coagulum forms and is removed by 
filtration through a folded paper filter. The filtrate 
contains the albumin dispersed in half-saturated 
ammonium sulphate solution. The salt can now be 
removed by dialysis in parchment, or, better, collo- 
dion against running water, and a moderately pure 
albumin sol obtained. The usual method is, how- 
ever, to precipitate the albumin by saturating the 
solution with ammonium sulphate, a further 38 to 
39 gm. being required for every 100 c.c. of filtrate. 
A thick coagulum of albumin forms, which is filtered 
overnight and allowed to drain as far as possible. 
The residue, which always contains a considerable 
amount of mother liquor, is then dissolved in the 
smallest volume of water which will give a clear sol, 
and dialysed as explained above to remove ammonium 

The same method is applied to fresh white of eggs. 


About 28 to 30 c.c. of white can be obtained from 
average fowls' eggs ; this contains about 10 per cent, 
of albumin. The total protein content is about 
12-2 per cent., the difference being accounted for by 
globulin and mucoid. The egg-white is beaten up 
with an equal volume of saturated ammonium sul- 
phate solution, which produces half-saturation in 
the mixture and precipitates the latter constituents. 
The coagulum is filtered off and the filtrate saturated 
with ammonium sulphate to precipitate the albumin, 
which is filtered off and dissolved in a small volume 
of water. This sol is again precipitated by satura- 
tion with ammonium sulphate, and the previous 
operations repeated ; several re -precipitations are 
required to obtain pure albumin. The last coagulum 
is dissolved in a small volume of water and dialysed 
to remove the sulphate. The losses are fairly con- 
siderable, and the albumin content of the sol finally 
obtained after dialysis will be between 2 and 2-5 gm. 
of albumin for every 30 c.c. of white used originally. 
Although the general student will hardly have 
occasion to use it, the classical method of making 
" crystallized albumin " is here given. It may be 
tried with the white of two or three eggs, which 
should be perfectly fresh. The volume of egg-white 
is measured and an exactly equal volume of saturated 
ammonium sulphate solution added to it in small 
portions at a time, the mixture being vigorously 
beaten with an egg-beater after each addition until 
the whole has been reduced to a stiff froth. This is 
allowed to stand overnight, and is then filtered to 
remove the coagulum of globuhn, etc. Ten per cent, 
acetic acid, i.e., glacial acetic acid diluted to 10 
times its volume, is then added to the filtrate from 
a burette, a single drop at a time, with gentle stirring 
to re-dissolve the precipitate formed locally before a 
further drop is added. This is continued until the 
solution becomes -permanently turbid — the exact 


degree of turbidity can only be found by practice, 
but must amount to something more than mere 
opalescence. When this point has been reached, 
I c.c. of acid for every loo c.c of solution is added. A 
copious precipitate forms, which, on standing and 
occasional gentle shaking, becomes (micro-) crystal- 
line after five or six hours ; to obtain the full yield 
it should, however, be allowed to stand for 24 hours. 
The precipitate is filtered off and dissolved in a small 
volume of water ; the solution is then dialysed, or, 
if further purification is desired, it is again preci- 
pitated. This is done by dissolving the coagulum 
from the filter in the smallest possible volume of 
water, acidifying with a few drops of 10 per cent, 
acetic acid, and then adding concentrated ammonium 
sulphate until a slight permanent turbidity results. 
After 24 hours' standing the bulk of albumin has been 

The beginner wiU find the exact degree of turbidity 
required somewhat difficult to judge and must be 
prepared for disappointment. In dialysing albumin 
sols remember what has been said on page 27 regard- 
ing sols which exert an appreciable osmotic pressure 
and use suitable arrangements. 


This is too voluminous to aUow of being summarised. 
Students must consult the text-books on Proteins or 
those of Biochemistry, 

Chapter VIII. 


Both as regards methods of preparation and pro- 
perties these faU into two classes, which are be=t 
studied separately : the pure oil-water emulsions, in 
which no solute is present in the water, and the con- 
centrated emulsions, which can be produced only by 
adding to the water phase certain substances which 
greatly lower its surface tension and occasionally 
possess other properties as weU. 

Pure Oil-Water Emulsions. — These are most con- 
veniently prepared by the following method : O'l c.c. 
of the oil (which may be a paraffin oil of low vis- 
cosity, oleic acid, or generally any other hquid 
immiscible with water, but soluble in alcohol) is 
dissolved in lo c.c. of alcohol or acetone. This 
solution is blown from a pipette into one litre of 
water ; the water is well agitated before immersing 
the pipette, the point of which should be lo to 15 cm. 
below the surface. The resulting emulsion should 
show a bluish tinge in reflected hght (particularly 
weU marked with oleic acid), and be practically clear 
in transmitted Hght. 

The emulsion should be examined with a dark- 
ground condenser and the sign of the charge deter- 
mined in the cataphoresis apparatus. The coagula- 
tion by HCl should be watched under the microscope 
as the phenomenon, viz., coalescence of discharged 
pcirticles to bigger ones, with decreasing amplitude 
of Brownian movement, is slow and more easily 
followed than with suspensoids, with the behaviour 
of which it otherwise agrees. 


Electrolyte coagulation should be tried with HCl 
and with, say, CaCla and Al2(S04)3 ; salts of univa- 
lent cations act only in very great concentrations. 
The effect of the coagulant shows itself macro- 
scopically by the disappearance of the bluish opales- 
cence, the emulsion becoming whitish and turbid 
instead. Samples should be taken at intervals and 
examined microscopically (ordinary illumination, 
using sub-stage condenser and a fairly small 
diaphragm, magnification about 600 diameters), 
when it will be found that globules about 3/i, diameter 
gradually take the place of smaller ones, this being 
the size at which Brownian movement becomes so 
sluggish that further colHsions between globules, 
and therefore formation of larger ones, practically 
cease to occur. 

The emulsions to which sufficient coagulant has 
been added gradually clear from the bottom upwards, 
provided the " oil " has a density lower than that of 
water. The rate of clearing should be measured at 
convenient intervals, 24 or 48 hours, according to the 
difference in density, and the size of the globules 
calculated from Stokes's formula [determine density 
of oil to three decimals). 

Concentrated Emulsions. — To prepare these it is 
necessary to lower the surface tension of the aqueous 
phase, the most convenient agent for the purpose 
being a soap. Either a soap solution may be used, 
the preparation of which will be described below, or 
the soap may be actually produced in the process of 
emulsification. This method, which of course is 
apphcable only to oils which are glycerides, consists 
in shaking up the oil with a dilute solution of caustic 
soda, N/50 to N/ioo being suitable concentrations. 
Small quantities may be prepared in test tubes ; 
pour 10 c.c. of the NaOH solution into a test tube of 
25 to 30 c.c. capacity,* then add ordinary olive or 
cotton-seed oil in lots of i c.c, close the test tube 


with the thumb and shake vigorously after each 
addition. The emulsion becomes a pure white 
{why P), and after the addition of about 10 c.c. of oil 
the viscosity increases so much that the dispersion of 
further oil becomes difficult. Larger quantities may 
be prepared in the same way in any shaking apparatus 
which may be available ; in this case, too, the oil 
should be gradually added in small portions. 

The oil in emulsions thus prepared gradually rises, 
a sharp boundary forming between the concentrated 
emulsion at the top and the dispersion medium, 
which is turbid owing to the presence of soap and 
very fine particles. This rise continues until the oil 
globules are in closest packing ; as they are not of 
uniform size, no exact figure can be given for the 
percentage of disperse phase, but it will be found 
to be 70 per cent, or over. The volume ratio can 
be determined, with small errors due to contrac- 
tion, etc., in the following way : a burette is filled to 
the lowest mark {i.e., the one bearing the highest 
number) with dilute hydrochloric acid. The emul- 
sion is then poured into the burette, the volume 
noted, the burette closed with the thumb, and emul- 
sion and acid thoroughly mixed. The oil separates 
and rises ; any small globules which may remain sepa - 
rated from the main bulk must be made to unite 
with it by tapping and inchning the burette. The 
volume of oil is then read off and the volume of 
continuous phase obtained by difference. 

When mineral oils are to be emulsified the proce- 
dure described is not apphcable, but soap solution 
must be employed. Ammonium oleate is extremely 
efficacious, but is not obtainable commercially, and 
sodium oleate (ohve oil or Marseilles soap) will 
generally have to be used. It is cut into fine 
shavings, which are allowed to dry in air for three or 
four days, and 10 gm. of the air-dried material 
dissolved in one litre of distilled water, at 30° to 



Fig. io. 

40° C. The solution is allowed to 
stand in the cold for 24 hours and 
then filtered twice through the same 
filter of fairly open paper. 

Emulsification may be accom- 
plished by shaking, as described 
above. If no shaking device is 
available, fair quantities may be 
prepared in the apparatus shown in 
Fig. 10, which may be made up 
from vessels to be found in the 
laboratory. A tall cyHnder is 
closed by a rubber stopper with 
two perforations. A thistle funnel, 
having a tube 3 to 4 mm. diameter, 
reaching to within 3 or 4 mm. of 
the bottom of the cylinder, passes 
through one of the perforations ; a 
ball tube, with two or three balls, 
through the other. A large pipette 
— 50 to 100 c.c. — with its lower 
end drawn to a capillary point, is 
suspended above the thistle funnel 
so that the point touches the waU 
of the funnel. The point is made 
so fine that 50 c.c. of oil of low 
viscosity, like the paraffins used in 
lamps, takes 25 to 30 minutes to 
empty. If the flow is found too 
rapid it may be reduced by fitting 
a short length of rubber tubing, 
provided with a screw clip, to the 
upper end of the pipette. 

The apparatus is used as follows : 
The stopper is removed and a 
known volume of soap solution 
poured into the cylinder ; it should 
not exceed one-third of the total 


volume. A few drops of the oil to be emulsified 
are then poured down the funnel and the latter 
rotated slowly, so that the whole of the tube is 
wetted by oil. The stopper is then replaced, the 
pipette filled with oil suspended as explained above, 
and the ball tube connected to the filter pump. 
The latter should be so adjusted that the air issues 
in a uniform string of separate bubbles at the 
bottom of the funnel tube. When the pump is 
working properly, the clip at the top of the pipette is 
opened to the required extent, and the apparatus 
then requires no further attention. The oil, which 
runs down the tube in a very thin film, is broken up 
by the air bubbles when passing out at the lower 
edge of the tube, and thorough emulsification takes 
place. Frothing is rather marked at the beginning, 
but subsides after a Httle oil has been emulsified, and 
the ball tube prevents froth from being drawn into 
the suction tube to any extent. 

The emulsions made with soap solution separate a 
" cream," like those prepared with caustic soda, and 
the same methods may be used for determining the 
volume ratio. 

Emulsions which do not separate, whatever the 
volume ratio, can be made from oils which have the 
same density as the dispersion medium. The 
simplest way is to prepare suitable mixtures of either 
olive oil, cotton-seed oil, or lamp paraffin with 
carbon tetrachloride (density at 0° 1*632). The 
ratio may be approximately calculated from this 
and the density of the oil ; as, however, the co-effi- 
cients of expansion of the aqueous dispersion medium 
and the mixture of oil and CCI4 differ widely, exact 
equality can only be secured by experiment at a 
definite temperature. For this purpose the — approxi- 
mate — mixture and a smaU beaker filled with the 
soap solution are placed in the thermostat, a i c.c. 
pipette filled with the former, and slowly blown out 



under the surface of the soap solution, so that a 
single drop is formed, which can easily be detached 
from the pipette. According as this drop rises or 
sinks, more CCI4 or more oil is added to the mixture, 
until the drop remains practically stationary for a 
few minutes. Emulsions of this kind are especially 
suitable for viscosity measurements (which see). 


For recent papers, see article " Emulsions " in Second 
Report of British Association Committee on Colloid 
Chemistry, 1918, p. 20, 

Chapter IX. 

The name was given by H. Bechhold to a method 
of separating the disperse phase of sols from the 
dispersion medium by means of filtration under 
pressure through porous membranes impregnated 
with gels, the permeability of which may be varied 
within wide Hmits. As considerable pressures may 
have to be used metal apparatus is essential, which 
limits the applicability of the method to some 
extent. An apparatus suitable for sols containing 
a very small amount of sohd only (as is the case with 
most suspensoid sols) is illustrated in Fig. ii. The 
filtering membrane rests on a perforated metal disc a, 
which is clamped between the body h of the filter 
and the slightly conical bottom c and rests on six 
radial -ribs in the latter. The branch d, closed by a 
screw cap, serves for filling the filter, and the 
necessary pressure is generated by a bicycle tyre 
pump connected to the valve e. The joint between 
the vessel and the cover is made tight by a rubber 
ring cemented to the spigot on the former with marine 
glue or Chatterton's compound. To prepare the 
filter for use, the cover with the perforated metal 
plate is removed, the membrane placed carefully on 
the latter ; the cover is then replaced and the two 
nuts tightened. The filter is then charged through 
the large inlet, the cap replaced and tightened, and 
pressure generated by means of the bicycle pump. 
The filter should not be more than about half full, 
so as to leave a sufficient air space, as otherwise the 



pressure falls too rapidly and the apparatus requires 
continuous attention and pumping. The pressure 
to be used depends on the denseness of the mem- 
brane ; two or three atmospheres (30 to 45 lbs. per 
square inch) will generally be sufficient, but the 

Fig. II. 

apparatus should be strong enough for a maximum 
working pressure of five atmospheres, i.e., it should 
be tested with eight atmospheres, as this adds very 
little to the cost. The apparatus is supported on a 
tripod stand, and the beaker or flask for the filtrate 
placed below it. 

The membranes are made from circles of filter 
paper impregnated with either acetic acid collodion 


or gelatin. A hard filter paper, such as used for 
vacuum filtration on Buchner funnels, must be 
employed ; brands equivalent to Schleicher and 
Schuell's Nos. 575 or 602 are suitable. 

Collodion membranes are more convenient than 
gelatin ones ; their permeability, i.e., the average 
size of the particles which are just retained, varies 
with the concentration of the collodion. As it is 
difficult, without experience, to foretell what con- 
centration will answer in any given case, a range of 
filters should be prepared impregnated with, say, 
I. 2, 3, 5, 7 and 10 per cent, collodion, i.e., sols of 
collodion cotton containing i, 2, etc., gm. of cotton 
in 100 c.c. of sol, the solvent being glacial acetic acid. 
The preparation of these sols has been fuUy described 
under "Dialysis," p. 24. 

If a number of filters are prepared with sols of 
various concentrations, the latter should be marked 
in pencil on the disc before impregnation. The sol is 
poured into a small dish — a porcelain developing 
dish does very well, — a filter paper seized in a forceps 
near the edge, within the width which will eventually 
be covered by the rubber joint ring, and slowly 
immersed in the sol under a very acute angle with 
its surface. Care must be taken not to trap any air 
bubbles underneath the paper and to have it uni- 
formly penetrated by the sol ; this is readily seen by 
the paper becoming translucent, like oiled paper, 
while spots not penetrated by collodion remain 
opaque and white. The thoroughly impregnated 
disc is then slowly withdrawn from the liquid and 
held vertically above it to allow the excess to drain 
off ; it should be turned to and fro in its own plane, 
as otherwise a thick ridge is formed at the bottom, 
which may prevent a good airtight joint being made 
when the disc is clamped in the apparatus. The 
discs are then submerged in water, which is con- 
stantly changed until all acetic acid has been washed 

72 wo. OSTWALD'S 

out, and may then be kept indefinitely in water 
saturated with chloroform or camphor, to prevent 
the formation of mould. 

Gelatin sols containing from 2 to 10 per cent, may 
be used instead of collodion, the sols being prepared 
in the usual way. The vessel containing the eol 
during impregnation must be placed in a water bath, 
and the temperature chosen should be maintained 
constant and adhered to throughout, as otherwise 
filters made even with the same concentration will 
vary considerably. The discs are impregnated and 
drained as described, during which time the sol sets 
to gel ; they are then hardened in a cold solution of 
formaldehyde, 2 to 4 per cent., which is placed in a 
refrigerator for 24 hours. The discs are then rinsed 
in water and can be kept under water saturated with 
chloroform. // either collodion or gelatin ultra-filters 
are allowed to dry, even partially, they become useless. 

Two per cent, collodion filters should give a colour- 
less filtrate with Prussian blue sol (prepared as 
described on p. 34], while 3 to 4 per cent, filters 
should retain practically the whole of the disperse 
phase in ordinary gold sols (reduced by formaldehyde 
or tannin). 

The use of special apparatus and the consequent 
contact with metal can be avoided by adopting 
either of the following methods of making ultra- 
filters, which are due to Wo. Ostwald. The first one 
furnishes membranes suitable for use with the filter 
pump, while the second one produces " spontaneous " 
ultra-filters, i.e., membranes of sufficient permeability 
to allow liquid to pass simply by hydrostatic pressure. 

I. Filters for Use with Vacuum. — The collodion 
used has the following composition : collodion 
cotton, 2 gm. ; alcohol, 14 c.c. ; ether, 84 c.c. ; or 
the commercial " Collodion P.B." or " Collodion, 
methylated " may be used. 

For preparing conical funnels, a circle of ordinary 


rough filter paper is folded twice in the usual way 
and placed in a well-fitting, smooth glass funnel. 
To ensure perfect fit it may sometimes be advisable 
to make one fold onty at first, to place the paper in 
the funnel, and to make the second fold when good 
contact all roimd has been secured by careful 
smoothing. The filter thus prepared is filled with 
collodion up to the edge ; when this has penetrated 
the paper over the entire surface, the excess is 
emptied, the filter being slowly turned while this is 
being done. Turning is continued while the collo- 
dion dries superficially (the time required depends on 
the room temperature and the desired permeabihty, 
and may be from three to six minutes) ; when it no 
longer sticks to the finger on being touched lightly, 
the funnel is placed in distilled water. The filter is 
ready for use after about 15 minutes' immersion, but 
may be kept under water indefinitely provided 
chloroform or camphor is added. 

The same method may be adopted for preparing 
Buchner funnels, but in this case a joint must be 
made between the edge of the paper and the per- 
forated plate hefore impregnation, as otherwise the 
collodion gets under the paper. A solution of 2 gm. 
of white crepe rubber in 100 c.c. of petroleum ether 
is used for the purpose. The funnel is inclined under 
an angle of 45° and a few cubic centimetres of the 
rubber sol poured down the side, care being taken 
that it does not reach the perforations. By turning 
the funnel round the sol is distributed to form a band 
round the perforated area, and while this is still 
" tacky " the circle of filter paper is placed in posi- 
tion and squeezed down all round. After a few 
minutes' drying a second rubber band is produced in 
exactly the same way, and when this has dried the 
paper is impregnated with collodion, as described 
above. When the collodion has dried to nearly 
the required extent a third rubber joint is made. 


and the funnel placed in distilled water, as before 

2. Spontaneous Filters. — 'The collodion for these 
contains 4 gm. of collodion cotton to 12 c.c. of 
alcohol and 84 c.c. of ether. A circle of ordinary 
rough filter paper is folded twice in the usual way, 
placed in a well-fitting smooth funnel, and thoroughly 
wetted with distilled water. Any excess is poured off 
or allowed to drain through ; if a little remains in the 
point of the filter it must be removed with a twisted 
spill of filter paper. The funnel is then partly filled 
with collodion, which is spread uniformly by inclining 
and turning, the excess poured off, and the collodion 
allowed to dry for four to five minutes, after which 
a second layer of collodion is poured in the same 
way. When this has dried a few minutes the filter is 
ready for use. A properly made filter of this kind 
should give a colourless filtrate with a gold sol made 
by reduction with formaldehyde. 

Buchner funnels may likewise be used, no rubber 
joint being necessary in this case. The paper is 
thoroughly wetted with distilled water and placed 
flat on the perforated plate. Sufficient collodion is 
then poured on to cover the whole of the paper, and 
the excess poured off, leaving, however, a remnant 
of two or three cubic centimetres, which is carefully 
distributed round the edge of the paper by inclining 
the funnel about 45° and turning it continuously 
until the collodion no longer flows. A second lot of 
collodion is poured in exactly the same way after the 
first one has dried to the desired extent, and the filter 
is ready for use when this is sufficiently dry. Par- 
ticular attention must be paid to getting a sufficient 
rim of collodion round the edge of the paper, as this 
makes the joint and prevents the liquid from 
escaping underneath the paper. 

An extremely convenient method of making small 
ultra-filters consists in the use of the seamless 


extraction thimbles, which can be obtained in a 
variety of sizes. If apparatus for using them with 
vacuum is available, they may be impregnated dry, 
as described under (i) ; otherwise it is more con- 
venient to impregnate them wet and use them as 
spontaneous filters by placing them in a loosely- 
fitting cylindrical funnel. 


For sizes of particles retained, determination of 
diameter of pores, etc., see the original papers by H. 
Bechhold, Zeitschr. f. phys. Chem., 64, 328 (1908) ; KolL- 
Zeitschr., XL, 3 and 33 (1907) ; III., 226 (1908). 

Chapter X. 

The simplest and most sensitive method of show- 
ing the presence of disperse matter in a Hquid is 
examination by the Tyndall cone, i.e., a narrow 
beam of intense hght projected through the hquid 
and viewed at right angles to the direction of the 
axis of the beam. The liquid to be examined is 
placed in a prismatic cell ; the small cemented 
specimen cells supphed by most makers of apparatus 
are suitable. If possible, cells cemented with 
dichromate-gelatin should be chosen, as they can be 
used both for aqueous solutions and organic solvents. 
One face of the cell is covered with black velvet, or, 
better still, the ceU is placed in a small wooden box, 
provided with two circular openings (Fig. 12) at 
right angles to each other and Hned with black 
velvet or painted a dead black with Indian ink. The 
light from a small hand-regulated arc lamp, Nernst 
lamp with horseshoe filament, or a tungsten arc 
(" point o' light lamp "), is projected through the 
cell by means of a lens so placed that the focus faUs 
about the middle of the cell, opposite the second 
opening, through which the path of the beam is 
viewed. All light should be screened. If particles 
are present the beam is visible ; when the disperse 
phase is colourless {e.g., mastic suspension) the cone 
shows a bluish tinge ; when it is coloured the beam 
may show a colour different from that of the Hquid 
viewed in transmitted light [e.g., red gold sols, in 
which the cone shows green) . It must be remembered 



that even true solutions of substances of high mole- 
cular weight, e.g., cane sugar, show the cone, and that 
it is not entirely absent even in filtered distilled 
water. Water freshly ultra-filtered and collected 
with due precautions against contamination by dust 
is nearly " optically void," i.e., the cone is invisible. 

By placing a suitable analyser, say a Nicoll prism 
mounted in a collar permitting it to be rotated, in 
the opening A, the Hght emitted by the cone can be 
shown to be polarized ; if the Nicoll is rotated the 
intensity of the cone varies, becoming a minimum in 

Fig. 12. 

two positions of the prism at i8o° from each other. 
Complete extinction occurs only with a disperse 
phase consisting of a non-conductor (mastic sus- 
pension) . 

Ultra-microscopic and Dark- ground Examination. — 
The "sht" ultra -microscope and the cardioid con- 
denser require large arc lamps and special wiring, and 
are beyond the scope of this book ; those in a position 
to use them will obtain all necessary instructions 
from the makers. Real ultra-microscopic illumina- 
tion is provided by the Jentzsch ultra-condenser, a 
section through which, showing the path of the rays, 
is shown in Fig. 13. It has the advantage of requiring 



only a 4 to 5 ampere hand-regulated arc lamp, of 
holding a comparatively large volume of liquid, so 
that adsorption effects are minimized, and of being 
very easily centred. The optical part is cemented 
into a cylindrical metal casing, which is closed by a 
metal cover provided with bayonet joint and a 
central quartz window for observation. The ultra- 
condenser is placed on the stage of the microscope so 
that the spigot on its lower side fits the opening in 
the former — the ordinary condenser being, of course, 
removed and the plane mirror used for illumination. 

The dimensions of the 
ultra - condenser do not 
permit the use of objec- 
tives of shorter focal 
length than 6 mm. (or 
I"), but, as the images are 
not geometrical, there is 
no Hmit to the eyepiece 
magnification permissible, 
and the highest power 
eyepiece available may 
be used. 

The condenser is filled 

Fig. 13. 

with the liquid to be examined by means of the inlet 
and outlet branches provided on the cover, care being 
taken not to leave an air bubble at the top of the liquid 
under the quartz cover. The light from the arc lamp, 
which should be provided with a lens giving a nearly 
parallel beam, is then directed on the plane mirror, 
which it should fill completely and uniformly, and 
the condenser placed in position. The hght is now 
centred, using the |" objective and a low power eye- 
piece, by adjusting the mirror until the brightly 
illuminated spot is exactly in the centre of the field. 
The low-power eyepiece is then replaced by the 
highest power available ; Zeiss's No. 18 compen- 
sating eyepiece, or an equivalent, is suitable. The 


most highly illuminated layer will be easily found by 
focussing up and down, and, as this layer is at some 
distance from any boundary surface, the Brownian 
movement will be seen in great perfection. The 
large diffraction rings which appear and disappear 
round many particles indicate, of course, vertical 
movement out of the focal plane. 

Coagulation of sols can be very conveniently 
studied with this condenser by running in an elec- 
trolyte solution through one of the branches on the 
cover, or by adding a small amount of coagulant to 
the sol before it is fiUed into the apparatus. After use 
the condenser must be carefully washed with dis- 
tilled water and thoroughly dried with linen free 
from grease. This applies equally to the metal 
parts in contact with the Hquid. 

If the ultra-condenser is not available, the pre- 
sence of at least coarser ultra-microscopic particles 
can be detected by means of one of the numerous 
dark-ground condensers. Typical forms of this 
apparatus are the Zeiss " Paraboloid " condenser, 
the Reichert "Table" condenser, and the Jentzsch 
" Concentric " condenser, which latter, like the 
Jentzsch ultra-condenser, is now made in this 
country. The methods of using and centring are 
slightly different with the different types, and are 
generally adequately described in the makers' 
pamphlets. Since they all depend on total reflection 
at the cover glass, slides of the thickness prescribed 
by the makers must be used in all cases. Both 
slides and cover glasses must be carefully cleaned in 
the following manner. They are washed in hot 
dichromate-sulphuric acid mixture for five to ten 
minutes and then rinsed thoroughly with distilled 
water. The slides or cover glasses are then seized, 
one by one, \vith a spring forceps and, after draining 
off the bulk of the water, placed in strong alcohol, 
in which they are kept until required. Immediately 


before use the slide is withdrawn from the alcohol 
by seizing one corner with a spring forceps, and the 
adhering alcohol burnt off over a spirit lamp or 
Bunsen burner. As soon as the slide has cooled it 
is placed on the condenser, ample cedar oil being 
used, as explained in the descriptive pamphlets. 
The cover glasses may be treated in the same way. 
provided they do not crack too frequently ; if this 
should be the case, the alcohol may simply be 
evaporated at a sufficient distance from the flame to 
prevent ignition. Before the cover glass is made 
ready a large drop of sol, free from any air bubble, 
should be placed on the centre of the sHde, and the 
cover glass dropped gently on it immediately it has 
cooled. The essential point of the method described is 
that slides and cover glasses are not touched with the 
fingers or with any textile material, as this renders them 
entirely useless for ultra-microscopic work. 

The layer of hquid between cover glass and slide 
is, of course, of very shght depth, and careful focus- 
sing on its central portion is necessary to observe 
particles moving freely. Many particles wiU always 
be found, by suitable focussing, to have adhered to 
the two glass surfaces. Electrolyte coagulation can 
be observed, though in a somewhat rough fashion, 
by placing a drop of solution on the edge of the cover 
glass, so that it can diffuse into the bulk of hquid. 
The most convenient objects for becoming familiar 
with the use of the apparatus are comparatively 
coarse systems, especially mastic or gamboge sus- 


Full information on the various types of condensers is 
to be found in the pamphlets issued by the makers 
(Zeiss, Leitz and Chas. Baker). Photometric investiga- 
tion on TyndaU cone, connection between size of particles 
and luminosity, etc., by W. Mecklenburg, KoU.-Zeitschr., 
XIV., 172 (1914) ; XV., 149 (1914) ; XVL, 97 (1915). 

Chapter XI. 

A SIMPLE apparatus, suitable for practice and pre- 
liminary work, may oe made up from glass parts 
available in every laboratory in the manner illus- 
trated in Fig. 14. A U-tube, about 250 mm. long, 
provided with an inlet tube at the lowest point of 
the bend, is supported in a suitable stand. The 
inlet tube is connected by a rubber tube, about 
350 mm. long, to a funnel capable of holding about 
75 c.c. The tube is fitted with a screw chp or with 
one of the patent chps provided with a catch, which 
allows it to be left fully opened, near the end of the 
inlet tube. 

Two electrodes are inserted in the tops of the limbs, 
consisting of foil rolled into a cyhnder, the diameter 
of which should be about 2 mm. less than that of 
the tube. Platinum is, of course, the best material, 
failing which, silver ; even cyhndrical sohd carbon 
electrodes may be employed, but small particles are 
liable to become detached from the latter during 
use. The electrodes are fixed to stout wires, which 
are best mounted in a strip of ebonite, acting also as 
distance piece, and provided with terminals. 

The apparatus is charged in the following manner. 
The clip is opened and the sol to be examined poured 
into the funnel, the latter being held so that its edge 
is about 10 mm. above the bottom of the bend. The 
liquid should just reach the latter ; the funnel is 
then lowered and again raised to the original level, 
to drive out any air which may have been trapped 

L.U. 6 



in the rubber tube, and the cock closed, with the sol 
standing just at the bottom of the U-tube. Distilled 
water is now poured into the latter so as to fill the 
limbs to about half their height. The funnel is then 

raised until the level oj 
the liquid in it is about 

1 mm. below that of the . 
water in the limbs, and 
the cock opened full bore. 
The funnel is now very 
slowly raised, the sol 
flows into the U-tube, 
and the level of the 
water in the hmbs rises 
correspondingly. The 
funnel must be raised 
at the same rate, i.e., 
the level of the sol in 
the funnel should never 
be more than i or 

2 mm. above the water 
level in the hmbs. If 
this is done properly, 
the sol rises in the U 
without mixing with 
the water, and even- 
tually shows a sharp 
boundary in both hmbs. 
If the funnel is raised 
too rapidly, or if the 
cock is not fully opened, 
the sol issues in a jet, 
impinges on the upper 

wall of the U-bend, and rises unequally in the limbs, 
without fining them completely and without forming 
the necessary sharp boundary surface against the 
water. While it is quite easy to fill the tube properly 
after a little practice, the beginner will find the pro- 

FlG. 14. 


cedure much facilitated by a loose plug of carefully 
washed cotton or glass wool, placed in the inlet tube 
at its junction with the bend. This checks and dis- 
tributes the admission of the sol and prevents its 
issuing in a jet. 

Sufficient sol must, of course, be admitted to raise 
the water level in the Umbs so far that the electrodes 
are covered completely. When this is the case they 
may be connected to the electric supply ; the hght- 
ing supply may be used, of course with sufficiently 
insulated wires or flexible leads. With a voltage of 
200 the gradient in a tube of the dimensions described 
above is about 5 V/cm. , so that a very distinct shifting 
of the boundary is noticeable after 10 minutes. The 
polarity of the electrodes must, of course, be deter- 
mined, e.g., with the " pole finding paper " obtain- 
able for this purpose. Rough measurements of the 
rate of travel may be made by marking the original 
boundary and measuring the displacement (choosing 
whichever boundary is the sharper) after a definite 
time. The field strength is the voltage divided by 
the distance of the electrodes ; the latter is somewhat 
uncertain, but the total distance in the axis of the 
tube, i.e., twice the length of the straight hmb, from 
the lower edge of the electrode, plus half the arith- 
metical mean of the internal and external circum- 
ference of the bend, may be taken as approximately 

For exact work a specially made apparatus is 
preferable, which incidentally avoids the use of 
rubber connections. A convenient form (after 
W. Nernst and A. Coehn) is illustrated in Fig. 15. 
It consists of a U-tube provided with two large 
cocks at the junctions of the straight hmbs with the 
bend. An inlet tube, about 3 or 4 mm. diameter, 
leads into the lowest point of the latter, and is bent 
at right angles to the plane of the dramng, ter- 
minating at the top in a charging funnel of suitable 






capacity. A scale of millimetres {not cubic centi- 
metres !) may be etched on the hmbs. The limbs are 
provided with electrodes such as described above. 

The apparatus is 
charged as follows. 
The funnel is filled 
with the sol to be 
examined, the cocks 
A and A' opened, and 
then the cock B, until 
the sol just rises 
above the large cocks. 
B, and subsequently 
A and A' also, are 
closed, and the small 
amount of sol in the 
limbs removed with 
spills of filter paper. 
The hmbs are then 
filled to the same 
height with distilled 
water and the elec- 
trodes placed in posi- 
tion. The cocks A 
and A' are now 
opened, and then B, 
which must be done 
I very slowly and uni- 
k / formly. The sol and 
\J the supernatant 
water in the Hmbs 
now rise ; sol is 
admitted until the 
electrodes are sub- 
merged, when they 
may be connected to 
the supply. What 
Fig. 15. has previously been 

f\ I I 


said regarding the electric gradient, of course, applies 
equally to this form of apparatus. 

If measurements of the velocity of cataphoresis 
are made, the results are usually reduced to unit 
gradient, as stated. To give an example, the 
boundary travels to the anode, the displacement 
amounting to 22 mm. in 10 minutes. The velocity 
per second is accordingly 2-2/600 cm. = 0-00366 
= 366 X 10-5. With a voltage of 240 V and a 
distance of 24 cm. between the electrodes the 
gradient is 240/24 = 10 V/cm. The velocity reduced 
to unit gradient is, therefore, 366/10 X lO"^ = 
36 X 10"^, which is a normal value for the more 
highly dispersed gold sols. 

Microscopic Observation and Measurement of Cata- 
phoresis. — Any one of the " dark-ground " con- 
densers described in the chapter on optical methods 
of examination may be employed. A shde must be 
provided with electrodes having parallel edges, 
between which the liquid is contained. The elec- 
trodes are two strips of metal foil ; platinum is, of 
course, preferable, faihng which silver may be used. 
The strips should be 3 mm. wdde and about 35 to 
40 mm. long for the standard microscope shde, 
i" X 3". They are fastened parallel to each other 
and at right angles to the length of the shde, so that 
the distance between the outside edges is equal to 
the width of the cover glass to be used. A I" cover 
will be found convenient, this making the distance 
between the outside edges of the strips about 22 mm., 
and the distance between the inside edges, i.e., the 
distance between electrodes, about 16 mm. 

The strips are fastened to the shde, which has 
previously been thoroughly cleaned in the manner 
already described, with Chatterton's compound (a 
preparation made for insulating purposes) . A small 
piece of the preparation is warmed sufficiently and 
drawn out into a thin filament about i mm. diameter. 


The slide and the two strips of foil are then placed on 
a warm metal surface and a piece of the filament 
about 25 mm. long laid on each strip, centrally as 
regards width, and at one end of the strip, so as to 
leave a length of 10 to 15 mm. clear. The strips, 
as soon as the filaments of compound have softened, 
are picked up with a forceps, inverted and placed 
parallel to each other on the slide the requisite dis- 
tance apart (the position of the outside edges can 
be previously marked with a diamond or a drawing- 
pen) . They should be dropped down into the correct 
position without subsequent shifting, to avoid 
smearing the slide with the compound. As soon as 
the strips are in position the slide is removed from 
the warm surface and placed on a clean piece of 
filter paper resting on a wood or glass table ; the 
strips are then weighted by placing them on a second 
microscope slide and on this a 50-gm. weight. 
Sufficient time — from 10 to 20 minutes according to 
the temperature of the room — must be allowed for 
the complete hardening of the cement. 

Electric connection to the electrodes may be made 
simply by pressing the wires down on the projecting 
ends of the electrodes, of course taking care to 
insulate the latter from the stage of the microscope 
by a piece of thin sheet rubber or the like. For 
repeated use it is, however, more convenient to solder 
leads of thin flexible cord to the ends of electrodes, 
the opposite ends being provided with terminals kept 
apart by a distance piece of ebonite. The flexible 
cords should be of sufficient length to allow the 
terminals to rest on the table, out of the way of the 
mirror, when the slide is in position on the stage of 
the microscope. 

The slide is placed on the stage and optical con- 
nection made with the particular condenser used in 
the manner prescribed for it. The slide should, if 
possible, be clamped down to avoid accidental 


shifting. A large drop of the hquid is then placed 
■ n the centre of the space between the electrodes ; 
this should be of sufficient size to cover the field 
between the edges of the two strips, and make good 
contact with both, when the cover glass is put on. 
The microscope is then focussed on the central layer 
of Hquid, so as to observe particles which are moving 
freely, and the current turned on. In view of the 
short distance between the electrodes a supply at 
4 to 5 V is sufficient (say, from accumulators or dry 
cells). The velocity may be measured by means of 
an eyepiece micrometer and stop-watch. Release 
the latter when a particle under observation passes 
through one of the numbered divisions, and arrest 
the watch when it has travelled, say, through five 
divisions. The actual value of the micrometer 
reading must, of course, be known, or determined 
in the usual way with a stage micrometer. The 
potential gradient is : voltage at terminals/distance 
of electrodes in centimetres. To obtain the velocity 
of the particles in unit gradient, i.e., one volt per 
centimetre, divide the velocity actually found by 
the gradient. For instance, a particle is found to 
travel five divisions of the scale in 32 seconds. Five 
divisions, with the particular objective and eyepiece 
used, correspond to, say, 0-24 mm. = 0-024 cm. = 
24 X io~^ cm. The velocity per second is, there- 
fore, 24/32 X 10-3 = 75 X 10-5 cm. With a vol- 
tage of four and a distance between the electrodes of 
1-6 cm., the gradient is 4/1-6 = 2-5 V/cm. To 
obtain the velocity in unit gradient, the figure found 
above must, therefore, be divided by 2-5, so that we 
finally obtain the velocity per second in a field of 
one volt per centimetre = 75/2-5 X lO"^ = 30 X 
io~^ cm. This is a normal value for a metal sol. 

The microscopic method is particularly convenient 
and rapid for determining the sign of the electric 
charge when only small quantities of the liquid 


under examination are available. When using it 
for this purpose, remember that the image is reversed, 
so that, if the anode is on the right, negatively 
charged particles will travel to the left in the field 
of vision. Measurements, however, may be subject 
to fairly considerable errors unless a layer of par- 
ticles is observed which moves freely and beyond 
the influence of the two glass surfaces. A cell 
designed to minimize the sources of error has recently 
been described by Th. Svedberg, a reference to which 
will be found in the literature given below. 


For the microscopic method generally, see A. Cotton, 
and H. Mouton, " Les ultramicroscopes, etc.," Paris 
(1906). For elimination of errors, Th. Svedberg and 
H. Andersen, Koll.-Zeiischr., XXIV., 155 (1919). 

Chapter XII. 


Before starting experimental work the student 
should commit to memor^^ a few t3'pical figures for 
the concentrations of uni-, bi- and tri-valent ions 
which produce precipitation in suspensoid sols. 
The following are representative : — 


Sign of 

Precipitation concin'ration in miHimoles/litre. 

AS2S3 . 



f J 
f 1 


NaCl 51-0 CaCli 0-65 AlCl, o-og 
NaCl 2-5 BaClj o-o5 AICI3 o-oi 
NaCl 1,000 CaCl.2 25 AlClj 0-2 
NaCl g-25 K.2SO4 0-20 — — 

These figures give the concentration existing in 
the mixture of sol and electrolyte ; if, e.g., i8 c.c. of 
the AsjSg sol is to be coagulated by 2 c.c. of NaCl 
solution, the latter wiU have to contain 510 miUimoles 
of NaCl, since by the addition of the sol it is diluted 
to one-tenth of its original concentration. 

A second point to be noted is the difference in the 
corresponding values, say, for NaCl, between the 
three negative sols. This may be specific, i.e., a 
given sol may be much more stable to electrolytes 
generally than others ; such is the case with the 
mastic suspension according to different observers. 
The precipitation concentration, however, also 
depends on the concentration of the sol, and this partly 


explains the difference between the AsgSg and the 
Pt sol, the former being much more concentrated. 

The figures for the AsoSg sol may be taken as 
typical for this and for the (dilute) Prussian blue 
sol, as described, while the figures for the Pt sol will 
be found approximately correct for the gold sol 
made by the formaldehyde method. The Prussian 
blue sol will be found the most convenient for the 
first experiments. 

The next point to be considered is the method of 
adding the electrolyte solution to the sol. A number 
of the classical investigations were carried out by 
titration, i.e., by adding electrolyte solution to the 
sol until perceptible coagulation took place. This 
method has the drawbacks that, unless there is a 
very marked colour change, as with red gold sols, it 
is by no means easy to notice an exact end-point, 
that it excludes the time factor, and that the con- 
centration of sol varies with different amounts of 
coagulating solution. Nevertheless the method 
gives a rough idea, and may be used for preliminary 
trials, care being taken to use fairly large volumes 
of sol and to place the beaker containing it during 
titration so that small changes in colour or turbidity 
can be readily noticed. 

The procedure to be adopted for exact deter- 
minations is as follows : a uniform volume of sol is 
fixed upon, to which is added a definite fraction of 
coagulating solution, the concentration of which is 
varied. In this way the sol concentration is kept 
uniform. The sol and solution are mixed by a 
uniform procedure, say closing the test tube con- 
taining it and reversing it twice or four times ; the 
mixture is allowed to stand for a definite time, say 
two or three hours, and is then examined. 

Eighteen c.c of sol and 2 c.c. of solution will be 
found convenient, the latter being diluted by the sol 
to one-tenth of its original concentration. As about 


51 miUimoles of NaCl, 0-65 of CaCla, and 0-09 of 
AICI3 respectively are the concentrations required 
in the mixture to produce precipitation, the concen- 
trations of the solutions used will have to be ten times 
greater, viz., 510 millimoles of NaCl, 6*5 of CaCl2, 
and 0-9 of AICI3. 

Since these concentrations just produce precipita- 
tion in a certain sol, it will be desirable to have a 
considerable margin, and the following three stan- 
dards for negative sols should be prepared : — 

NaCl . 1,000 millimoles = 58-5 gm. in a htre. 
CaCla .15 ,> = 1-665 gm. 

AICI3 . 2 „ = 0-267 gm. 

For very accurate work, or if the materials are of 
doubtful purity, these solutions should be stan- 
dardized against suitable standard solutions. 

Since one part of these solutions added to nine 
parts of sol will certainly produce coagulation, more 
dilute solutions of known strength must be prepared. 
Do this by mixing in test tubes, say, 8 c.c. of solution 
with 2 c.c. of water, 6 c.c. of solution with 4 c.c. 
of water, 4 c.c. of solution with 6 c.c. of water, 
and 2 c.c. of solution with 8 c.c. of water ; label 
the tubes o-8, o-6, 0-4 and 0-2, these being their 
respective concentrations referred to the stock 

Now place 18 c.c. of the sol to be examined in each 
of five test tubes, label them i, o-8, o-6, 0-4 and 0-2, 
add to each 2 c.c. of the corresponding solution, mix 
by the standard method decided upon, and allow 
the tubes to stand for a definite time, say two or 
three hours. It will then be found, e.g., that the 
contents of i, o-8 and o-6 have been precipitated, 
but that those of 0*4 and 0*2 have not changed. The 
hmit concentration accordingly lies between the 
electrolyte concentrations prevaiUng in o-6 and 0-4. 
The solutions added, if NaCl was used, contained 


1,000 X 0-6 and i,ooo X 0-4, i.e., 600 and 400 milli- 
moles respectively ; since they were diluted to one- 
tenth by the sol, the actual concentrations in the 
mixture are 60 and 40 millimoles. The minimum 
concentration necessary for precipitation lies between 
these two ; to determine it more accurately, an 
intermediate concentration of the added solution 
may now be tried, say 0-5 of the original. For this 
purpose mix, say, 2 c.c. of stock solution with 2 c.c. 
of water, and add 2 c.c. of the mixture to 18 c.c. of 
sol as before. If precipitation just occurs within the 
standard time, the Umit concentration is obviously 
1,000 X 0-5/10 = 50 millimoles per litre. 

The procedure just described should be carried out 
with several sols, e.g., Prussian blue, arsenic tri- 
sulphide and gold reduced by formaldehyde, with all 
three electrolytes, and the results tabulated. The 
results should be compared with the numerous data 
given in the literature and carefully checked if they 
show very marked deviations from the average. 
The contents of the sols in disperse phase should also, 
for comparison, be calculated from the data given 
for their preparation. 

A similar procedure should be adopted with the 
ferric hydroxide sol, the only representative of the 
positively charged sols. By reference to the table at 
the beginning of the chapter we find that the Umit 
concentrations are 9-25 miUimoles of NaCl and 
0-20 of K2SO4 (Na2S04 may be used instead). 
Using the same ratio as before, 2 c.c. of solution to 
18 c.c. of sol, the concentrations of the former wiU be 
92-5 and 2-0 milHmoles, and to have the same margin 
as before we shall require stock solutions of the 
following concentrations : — 

NaCl . 200 millimoles = 117 gm. in a Utre. 
Na2S04 . 4 ,, = 0-568 gm. „ 

The NaCl solution may, of coiirse, be made up from 


that previously used for negative sols by diluting 
with four volumes of water. 


For electrolyte precipitation generally, W. D. Ban- 
croft, Second Report of British Association Committee 
on CoUoid Chemistry, 1918, p. 2. A very complete study 
of fractional precipitation in Sven Oden, " Der koUoide 
Schwefel," N. A. Reg. Soc. Scient. Upsal., Ser. IV., 
Vol. 3, No. 4 (1913). Coagulation velocity : H. H. 
Paine, Koll.-Zeitschr., XL, 115 (1912) ; H. Freundlich, 
and C. Ishizake, Faraday Soc. Gen. Discussion on 
Colloids and their Viscosity, 1913 ; H. R. Kruyt and J. 
van der Speck, Koll.-Zeitschr., XXV., i (1919), a very 
careful study of electrolyte coagulation. 

Chapter XIII. 



Sols in which the disperse phases carry opposite 
charges precipitate each other when mixed in 
definite ratios, while no precipitation occurs if an 
excess of either sol is present. 

The ferric hydroxide sol described above will, 
generally speaking, precipitate an equal volume of 
the Prussian blue, the gold sol reduced by formalde- 
hyde, or the mastic suspension. Place 5 c.c. of each 
of the negative sols into test tubes, add to each 5 c.c. 
of the (dialysed !) ferric hydroxide sol, mix by a 
uniform procedure, and allow the tubes to stand. 
The Prussian blue and the gold sol will generally 
show coagulation within a few minutes, while the 
mastic suspension may take 15 to 25 minutes. The 
coagulum contains both disperse -phases, so that the 
liquid in the test tubes is colourless after the former 
has settled. 

If precipitation fails to occur with any of the sols 
mentioned, or with any other negative sol mixed 
with an equal volume of ferric hydroxide sol, the 
correct ratio must be ascertained by experiment. 
For this purpose place in test tubes i, 2, 3, etc., up 
to 9 c.c. of ferric hydroxide sol and add (in the same 
order) 9, 8, 7, etc., down to i c.c. of the negative sol. 
The contents of each tube must be mixed, by a 
uniform procedure, immediately after the second sol 
has been added. After, say, one hour note the ratio 
in the tube or tubes in which coagulation has occurred. 


Examine the electrical condition of two mixtures 
in which no coagulation has occurred, one having 
ferric hydroxide, and the other negative sol in excess. 
For this purpose note the ratios and then make up 
a sufficient quantity of the mixtures for cataphoresis 
in the U-tube. The sign of the charge in the mixture 
will be found to be that of the sol present in excess, 
the charge on the other sol having been reversed. 

Chapter XIV. 


The protective effect of emulsoids may be demon- 
strated in two ways. The emulsoid may be added 
to one component of a reaction which produces a 
precipitate and may cause the latter to become much 
more highly disperse than it would be in a pure 
aqueous medium. Or the emulsoid may be added to 
an existing sol, in which case it protects it from electro- 
lytes, i.e., the concentration of the latter necessary 
to produce coagulation is considerably increased. 

To demonstrate the formation of a highly disperse 
precipitate in the presence of a protective colloid, 
dissolve 0*5 gm. of crystallized barium chloride in 
50 c.c. of water, and 0-25 gm. of ammonium sulphate 
in 50 c.c. of water. Add 5 c.c. of the first solution to 
5 c.c. of the second, and note that the bulk of the 
precipitate settles in a few minutes. 

Warm the ammonium sulphate solution to about 
30° C. and add to it 5 c.c. of 15 per cent, gelatin sol, 
mix thoroughly, and then add the barium chloride 
solution with continual stirring. The precipitate 
does not settle out on standing, and the mixture 
passes through a close filter paper without leaving 
any residue. Many other precipitates may be 
obtained as sols in this way by adding to one 
solution varying amounts of gelatin, albumin or gum 
arable, and by choosing suitable concentrations. 

The other procedure is to add to a sol, for which 
the electrolyte concentration required to produce 
coagulation in a definite time has been previously 
determined, small amounts of gelatin, albumin or 
gum arable sol and to determine what concentration 


of electrolyte will now produce a marked change or 
rapid coagulation. Thus, 9 c.c. of the gold sol made 
by the formaldehyde method (p. 30) turns blue 
within a few seconds after the addition of i c.c. of 
N/i NaCl solution. (Watch this change by pure 
transmitted hght, say by looking through the test 
tube at a uniformly illuminated screen of white 
paper ; reflected Hght must be excluded, as the 
strong reddish-brown smface colour of the sol is 
almost the same for the original red as for the blue 
sol and makes observation of the change difficult.) 
Now add to 9 c.c. of the same sol i c.c. of a o-i per 
cent, gelatin sol (o-i gm. in 100 c.c), mix well, add 
I c.c. of the N/i NaCl solution, and note that no 
change of colour occurs, even on standing. Add more 
NaCl solution, i c.c. at a time, and note that even 
4 or 5 c.c. produces no change whatever. 

Gold Numbers. — The gold number of an emulsoid 
is defined by R. Zsigmondy as the number of milli- 
grammes of the emulsoid just sufficient to prevent a 
colour change in 10 c.c. of a standard red gold sol on 
addition of i c.c. of a standard solution of NaCl 
(density 1-07, i.e., concentration about N/i). As 
these figures are arbitrary, it is better to state the 
percentage concentration of emulsoid in the gold sol 
which just prevents the colour change when i c.c. of 
N/i NaCl solution is added to about 10 c.c. of sol. 
The following figures have been calculated from 
Zsigmondy's gold numbers : — 

Minimum concentration in 


per cent, which prevents 
colour change. 



. 0-00005 

to o-oooi 

Egg albumin 1 



mercial) . 


. O'OOI 

to 0-002 

Gum arable 


. 0-0015 

to 0-0025 



. o-io 

to 0-20 

Potato starch 


• 0-25 



These are exiguous concentrations, especially for 
the most active protective agents ; if i c.c. of emul- 
soid sol is to be added to lo c.c. of gold sol, the con- 
centrations of the former will have to be ii times 
those given in the table, e.g., 0-00055 to o-ooii per 
cent., or 0-0055 to o-oii gm. per litre, for gelatin. It 
is advisable to make up sols of still greater concen- 
tration, say 50 times those given in the table, 
preferably by suitable dilution of any concentrated 
sols available in the laboratory. 

Then proceed as follows : Place 10 c.c. of gold sol, 
which should be a pure red without any purple 
tinge, into each of a number of test tubes, add o-i, 
0-2, 0-4 up to I c.c. of emulsoid sol and mix. Then 
add to each test tube i c.c. of N/i NaCl and note the 
concentrations in the tube which just retains the red 
colour and the next one, which shows the change to 
blue. If, for instance, the tube with 0-2 c.c. of 
emulsoid sol remains unaltered and the one with 
o-i has turned, the emulsoid concentrations are 
respectively : — 

2/102 == 0-0196 X original emulsoid concentration, 
i/ioi = 0-0099 X 

The gold number in percentage hes between these 
values and may be determined more exactly by using 
intermediate volumes or by reducing the concen- 
tration of the emulsoid sol used and repeating the 

The gold numbers are a delicate means of differen- 
tiating between proteins which cannot readily be 
distinguished by other tests. The method has, 
therefore, acquired some importance in medical and 
biochemical work, which is, however, beyond the 
scope of this book. 

Behaviour of different Sols. — The protective effect 
of a given emulsoid sol is not necessarily the same on 
other sols as on gold sols ; apart from the question 


of concentration there appear to be specific differ- 
ences, although, generally speaking, the various 
emulsoids stand in the same order for most sols as 
they do for gold sols. To show this, determine 
approximtely the volume of N/i NaCl solution neces- 
sary to coagulate 10 c.c. of the (dilute) Prussian blue 
sol. Add I c.c. of the o-i per cent, gelatin sol pre- 
viously used with gold sol to 9 c.c. of Prussian blue 
sol, then add the volume of sodium chloride solution 
found sufficient to coagulate the unprotected sol. 
Generally this wiU be sufiicient to precipitate the 
protected sol after a somewhat longer time, not- 
withstanding the presence of an amount of gelatin 
which, in the previous experiment, completely pro- 
tected the gold sol. 


For protective effect and gold numbers, see Zsigmondy- 
Spear, " Colloidal Chemistry " (Chapman and Hall, 
1917), pp. 106 — III ; "Investigations on a Number of 
Protective Agents, chiefly of Vegetable Origin," by A. 
Gut bier and collaborators, Koll.-Zeitschr., XVIII., i, 
57, 141, 201 ; XIX., 22, 90, 177, 230 (1916) ; XX., 
123, 186 (1917). 


Chapter XV. 


The only suitable instrument for accurate deter- 
minations is a properly designed capillary viscometer. 
The various rough methods employed occasionally, 
such as determining the time required for a given 
volume to flow from a pipette, or the time taken by 
a small sphere to fall through a given height, are 
useless, as in all these arrangements the time 
measured is very far from being simply proportional 
to the viscosity. 

Two types of capillary viscometers may be used : 
the simple Ostwald type (Fig. i6), in which the flow 
is caused by the difference of head in the two limbs 
of the instrument, and the Ubbelohde type (Fig. 17), 
in which the flow is caused by a constant air pressure 
applied to one limb. The use of the latter and the 
manostatic apparatus for providing constant air 
pressure will be described later. 

The Ostwald viscometer consists of a wide tube, 
generally provided with a bulb at the lower end, 
which is joined by a bend to a straight capillary tube. 
The latter leads into a bulb capable of holding 2 to 
3 c.c. of liquid, and provided with an inlet tube, a 
constriction being provided where this tube joins the 
bulb. Marks are placed, one in the centre of the 
constriction and one below the bulb at the beginning 
of the straight capillary. The instrument is gene- 
rally used only for determining relative viscosities, 
e.g., the viscosity of a sol referred to the viscosity of 
the pure dispersion medium as unity. In this case 
all the factors in Poiseuille's formula which depend 



on the measurements of the instrument remain the 
same, and the pressures in the case of two hquids 




Fig. i6. 

Fig. 17. 

having the densities p^ and p^ are simply propor- 
tional to these densities, provided the same volume 
is used, so that the effective height of the liquid is 
the same in both cases. If the times taken by the 


level of the liquid in sinking from the upper to the 
lower mark are respectively t^ and-^i, the viscosities 
YJQ and Yji are in the following ratio : — 

and the relative viscosity rj-^ 

H Pq 

It is accordingly necessary to determine the 
density of the liquids under examination, and it need 
hardly be added that this must be done at the tem- 
perature or temperatures at which the viscosities are 
to he measured. 

To use the instrument it is first mounted vertically, 
by sighting it with a plumb-line in two directions 
at right angles to each other. A definite volume of 
Hquid is then run into the wide tube from a pipette 
reserved for this purpose. A convenient length of 
rubber tube is fitted to the top of the bulb and the 
liquid drawn up through the capillary until it has 
filled the upper bulb and risen well above the upper 
mark, when the tube is closed by pinching with the 
left hand. (Sufficient liquid must be delivered by 
the pipette to leave some liquid in the lower bulb 
when this has been done.) A stop-watch is held in 
the right hand, the rubber tube released and the 
watch started when the level of the hquid passes 
through the upper mark, and stopped when it passes 
through the lower mark. The eye should he kept on 
the viscometer and not on the watch. 

As viscometers are not readily obtainable, they 
will generally have to be blown specially, when the 
following points are to be noted. For sols with 
water as dispersion medium and, therefore, requiring 
the time of outflow of water to be determined, 
capillaries about 0-5 to o-6 mm. bore will be 
suitable. The length of the capillary should be 
80 to 100 times the diameter of the bore, say 5 to 


6 cm. at least. A bulb holding 2 to 2-5 c.c. on the 
end of the capillary is convenient, and the diameter 
of the inlet tube, as well as that of the bend connect- 
ing the capillary with the wide tube, should not be 
less than 3 mm., preferably 4 mm. The change in 
diameter from the bulb into the capillary, and from 
the latter into the bend, should be smooth and 
gradual. The time of flow should not be less than 
60 seconds for water at 20° C. 

Viscometers conforming to this description will be 
suitable for measuring viscosities up to 20 or 25 
times that of water, i.e., for fairly concentrated 
emulsoid sols. Sols having organic dispersion media, 
such as rubber-benzene sols, or sols of nitrocellulose 
in various media, however, often have viscosities of 
much larger order even in moderate concentrations, 
and it is then not feasible to carry out the whole 
series of measurements with the instrument used for 
the dispersion medium. A range of viscometers of 
increasing bore, and proportionately increasing 
length of capillary, must then be provided. The 
readings made by two instruments, say one used for 
the dispersion medium and for relative viscosities 
up to 20, and the next instrument, are then connected 
with each other in the following manner. The time 
of efflux for the most concentrated sol, for which the 
first instrument can be used, is determined and 
found to be Wj. The time of efflux for the same sol 
is then determined in the second instrument and 
found to be a smaller value, «2- To reduce readings 
on the second instrument to those on the first, and 
eventually to relative viscosities, they must, there- 
fore, be multiphed with n-^jn^- Instead of the 
sol itself any sufficiently viscous liquid may be used 
to determine this ratio ; glycerine or mixtures of 
glycerine with (httle) water may be employed. 

Viscometers must be thoroughly cleaned imme- 
diately before use with hot dichromate-sulphuric 


acid mixture, followed by distilled water and then 
by alcohol and ether, which is dried off by blowing 
air filtered through glass wool through the instru- 
ment. The latter precaution is necessary, as even 
small particles of dust may vitiate results seriously, 
in view of the small bore of the capillary. For the 
same reason the liquids to be investigated should be 
filtered, or, where this is not possible, at least 
strained through a glass wool plug. 

Since viscosity decreases to the extent of 3 to 
5 per cent, per degree of temperature in pure liquids, 
and at a much higher rate in emulsoid sols, it is 
absolutely essential that measurements should be 
carried out in a thermostat in which it is possible to 
keep temperatures constant within o-i° C. Where 
a proper apparatus is not available a beaker holding 
at least two litres may be used. The viscometer, a 
thermometer divided into tenths of a degree, and the 
toluene regulator are supported in the water from a 
suitable stand, and some form of stirrer must be 
arranged. The beaker stands on asbestos-coated 
wire gauze on a tripod, and is heated by a small — 
pin-hole or " micro " — gas burner, which must, 
however, be sufficient to keep the water at the 
required temperature, i.e., to make up the heat lost 
by convection and radiation. The water may be 
heated up to within a degree of the required tempera- 
ture by a Bunsen burner, and the small burner sub- 
stituted for it then. The regulator is then adjusted 
so as to cut off about the required temperature ; 
since it is somewhat troublesome to do this with 
complete accuracy, it is better to take readings 
within 0*2 or 0*3° of round numbers rather than 
to spend much time in trying to set the regulator 
exactly to the latter. Thus, if measurements at, 
say, five degrees' intervals are wanted, it will be quite 
permissible to work at 20-3°, 25-1°, 30-0°, etc., pro- 
vided the results are plotted accurately. 


The temperature regulation may fail in two ways : 
either the temperature may fall, although the regu- 
lator is fully open — in that case the gas pressure or the 
size of the burner is insufficient ; or the gas regulator 
may fail to cut off, although the temperature keeps 
rising — this trouble is generally due to " creeping " 
of the toluene, which passes between the mercury 
and the glass, instead of raising the mercury. This 
can be corrected by renewing the toluene and 
thoroughly cleaning the mercury. The trouble is 
much less frequent if a concentrated solution of 
calcium chloride in water is substituted for the 

The determination of the density — unless required 
for some further reason — is tedious and can be 
avoided by the use of the viscometer illustrated in 
Fig. 17, in which the pressure causing the flow is 
produced by compressed air instead of by the head 
of liquid itself. The instrument has two bulbs of 
equal size ; one limb of the U connecting the bulbs is 
a capillary of suitable bore, while the other is a wider 
tube. The hquid is drawn into the viscometer 
through A, and a definite volume must be used, so 
that, when the level is at B in the bulb on the 
capillary side, it stands at C in the opposite limb. 
When pressure is applied on the side containing the 
capillary, the Hquid rises into the opposite bulb, and, 
finally, the difference of levels is equal and opposite 
to that which prevailed at the beginning, so that the 
effect of the Hquid head, and, therefore, of the 
density, is eHminated. 

The pressure is generated by the simple manostat 
'shown in Fig. 18. A Mariotte's bottle A discharges 
water through a rubber tube and a piece of glass tube 
turned upwards at a right angle into a second bottle 
B of the same size. A Tee-piece passing through the 
stopper in the top of the bottle is connected at C to 
a water pressure gauge and at D to the viscometer, 


which is fitted with a three-way stop-cock. The 
active column of water is that between the bottom 
of the air tube of the Mariotte bottle and the top of 

Fig. i8. 

the bent tube discharging the water, and the gauge 
must, of course, have a limb of somewhat greater 
length than this height. 

Before using the apparatus, water is allowed to 
flow into the lower bottle until the column in the 


gauge becomes stationary. While this is being done 
the three-way cock is turned so as to shut off the 
compressed air tube from the viscometer and to 
leave the latter open to the atmosphere. The 
viscometer is now filled with the required volume of 
liquid, which is drawn up into the left-hand Umb 
above the bulb by suction apphed at E. The cock 
is then turned so as to shut off the viscometer from 
the atmosphere and leave the compressed air supply 
also shut off. The stop-watch is now got ready, the 
compressed air admitted to the viscometer, and the 
watch is released as the hquid passes through the 
lower mark and arrested when it passes through the 
upper mark. All measurements in one series are, 
of course, carried out with the same air pressure, 
i.e., without altering the column of the manostat. 
The time between marks is then directly proportional to 
the viscosity. 

It is hardly necessary to add that this type of 
instrument, like the Ostwald viscometer, must be 
kept in a thermostat. In both cases it is essential 
to make sure that the liquid under examination has 
reached the temperature indicated by the thermo- 
meter in the water bath ; this is the case if two 
readings taken at an interval of, say, five minutes do 
not differ by more than i per cent, at the outside. 
Generally speaking, three determinations on the 
same specimen should always be made and the 
arithmetical mean taken as the final result, provided 
the three readings do not differ by more than i per 
cent. If the readings decrease, the temperature 
may still be rising, as already pointed out ; if, how- 
ever, the decrease becomes more marked on repeti- 
tion, the viscosity of the sol is being reduced through 
its being forced through the capillary. This pheno- 
menon is quite common with many emulsoid sols 
and very marked, e.g., with starch sols. 

If the readings increase on repetition (provided, of 


course, that the thermostat is working properly), the 
capillary is becoming blocked, either by accidental 
contamination with dust, etc., or by adsorption or 
coagulation on its wall. In that case the instrument 
must be removed and thoroughly cleaned in the 
manner already described. 

In general series of viscosity measurements will be 
carried out to determine, e.g., the change in viscosity 
with concentration, or — concentration being con- 
stant — with temperature. In both cases the results 
should be plotted on sectional paper, to a fairly large 
scale, with the variable concentration (or tempera- 
ture) as abscissa and the viscosity as ordinate. The 
latter is, of course, proportional to the product : 
time X density, if the Ostwald viscometer is used, 
and to the time only with the Ubbelohde viscometer. 
The points found should, in general, lie very nearly 
on a smooth curve ; cusps or inflexions occur only 
where coagulation or the like takes place. Points 
which fall outside a smooth hyperbolic or logarithmic 
curve to a marked extent are, therefore, suspect 
unless such disturbing phenomena are at all pro- 
bable, and the particular reading should be carefully 

To practise the use of the apparatus the beginner 
will find gum arable sol convenient. Five concen- 
trations, say 5, 10, 15, 20 and 30 per cent., should be 
prepared, filtered through glass wool and the vis- 
cosities determined at some convenient temperature, 
say 2° or 3° C. above that of the room. The sols 
should be prepared when required. For viscosity- 
temperature measurements a sol made from 30 gm. 
of gelatin in 100 c.c. of water is convenient, made 
and filtered in the usual way. Measurements should 
be begun with the thermostat at, say, 45° or 50° ; 
the temperature is then allowed to fall about 5°, the 
regulator readjusted and a reading taken, and this 
procedure is continued until the setting temperature 


of the sol is reached. The results may be plotted 
directly, i.e., as the products time X density ; a 
clearer insight is, however, gained by plotting 
relative viscosities with water at the same temperature 
taken as unity. The viscosity of water at different 
temperatures may be found in tables, or be deter- 
mined in a second viscometer placed in the thermo- 
stat ; the densities of water at different tempera- 
tures are given in most works of reference. The 
quantities to be plotted will then be, if the times of 
efflux and the densities of the sol at the temperatures 
I, 2, etc., are t-^ and p-^, t^ and p^, etc., and the 
corresponding figures for water t\ and p\, t' 2 and p\. 

niln'i = h Pil^'i P'l ' V2h'2 = h P2/i'2 p'2. etc. 
A curve plotted with these relative viscosities as 
ordinates against temperatures as abscissae shows 
that the temperature coefficient of the sol is much 
greater than that of water, i.e., the percentage 
decrease with rising temperature is much greater 
than that of pure water. 


See, generally, Faraday See. Gen. Discussion on Colloids 
and their Viscosity, 1913 ; copious references to experi- 
mental work are given in this, especially in Wo. Ostwald's 

Chapter XVI. 


The removal of solutes from solution by solids 
having a large surface can be shown in a great variety 
of ways. Solutions of dyes, e.g., crystal violet, 
methyl violet, methyl green, etc., containing 2 to 

3 mg. in 100 c.c, may be shaken with 2 to 3 gm. 
of charcoal, fuller's earth or china clay, and will 
generally be found colourless after the adsorbent 
has settled. The adsorption of lead salts is another 
striking example. Add to 100 c.c. of water 2 or 3 
drops of concentrated solution of lead nitrate, take 
5 c.c. of the mixture and note the reaction with 
ammonium sulphide. Then shake the buUc with 

4 to 5 gm. of charcoal, filter and test the filtrate with 
ammonium sulphide ; with moderately good brands 
of charcoal no reaction, or at most a very faint 
brown tinge, will be visible. 

Influence of Solvent on Adsorption. — Dissolve 2 to 
3 mg. of methyl violet in 100 c.c. of water, shake 
with 2 gm. of charcoal and allow the latter to 
settle ; the supernatant liquid is generally colour- 
less. Pour it off as far as possible, and pour on the 
charcoal 80 to 90 c.c. of alcohol or acetone. This 
immediately assumes a violet colour, showing that 
the equilibrium concentration in the organic solvent 
is higher than in water, i.e., the amount^adsorbed 
from it is smaller. 

Adsorption due to Neutralization of Electric Charges. 
— ^A glass tube, about 25 mm. diameter and about 


60 cm. long, is held vertically in a suitable clamp. 
The lower end is closed by a rubber stopper, through 
which passes a short piece of glass tube about 4 mm. 
diameter, the upper end of which is flush with the 
surface of the stopper, while the lower projects 
I or 2 cm. Place a loose plug of glass wool on the 
stopper, and then fill the tube to half its height with 
silver sand, which has been washed with nitric acid 
followed by water, and has then been dried and 
ignited. Then fill the tube with ferric hydroxide sol 
and collect the hquid which escapes from the outlet 
tube in a beaker. This is quite colourless, the ferric 
hydroxide, which is positive, having been discharged 
and retained by the negatively charged quartz 
grains. If Night blue, a dye which is also positively 
charged in aqueous dispersion, can be obtained, the 
sol may be used instead of ferric hydroxide ; 2 to 
4 mg. in 100 c.c. is a suitable concentration. 

Similar results may be obtained by allowing strips 
of filter paper (which also takes a negative charge 
in water) to dip into sols. If the latter are positive, 
only water rises, the disperse phase being coagulated 
at the level of the liquid ; if the sol is negative, no 
separation occurs and the colour, e.g., of Prussian 
blue, rises in the strip. 

Selective Adsorption. — An example can be demon- 
strated as follows : Dissolve 5 gm. of gelatin in 
50 c.c. of water in the usual manner, and pour the 
sol into shallow moulds, so as to obtain strips or 
discs 3 to 4 mm. thick. Remove these from the 
moulds after 12 hours, and place them in a flat 
porcelain (developing) dish containing about 150 c.c. 
of a 2 per cent, solution of commercial aluminium 
sulphate. Place 10 c.c. of the solution in a test tube, 
add a few drops of ammonium thiocyanate, note 
that the solution shows a marked iron reaction, and 
set the sample aside. After lying in the solution 
for three or four days the gelatin shows a marked 


reddish-brown tinge due to ferric iron ; if a lo c.c. 
sample of the solution, in which the gelatin is lying, 
is again tested with thiocyanate and compared with 
the original sample, the iron content will be found 
to be much reduced. 

Chapter XVII. 


This is a method of separating and detecting 
various constituents of a mixture by means of the 
difference in their rates of diffusion and adsorption. 
The usual procedure is to allow the solution con- 
taining the several solutes to rise in strips of white 
filter paper ; the various constituents rise to different 
heights, and may be detected by their colour, or by 
suitable reagents applied to different portions of 
the strip. 

The strips should be cut from a white, neutral 
filter paper (Whatman No. 2 is suitable) about 
I cm. wide and 25 to 30 cm. long ; the edge of the 
sheet must not be used. They are then suspended 
vertically, with their lower ends dipping about 2 cm. 
into the hquid to be examined. Evaporation must 
be prevented ; for single experiments the simplest 
way is to place the liquid in the bottom of a tall 
cylinder and suspend the strip from the stopper, care 
being taken that it hangs \-ertically and does not 
touch the waU. The hquid is allowed to rise until 
it becomes stationary or for a fixed time, say 6 
to 9 hours, and the strip is then examined and 

The following is a convenient example for showing 
the dehcacy of the method. Shoe about 100 gm. 
of boiled beetroot, pulp the shoes with 50 c.c. of 
5 per cent, acetic acid, place the pulp into a musHn 
bag and express about 50 c.c. of liquid, which need 
not be filtered. Take 5 c.c. of the liquid and add to 


it gradually in a test tube N/g^ caustic soda solution ; 
the colour changes to purple, brown and, finally, to 
a dirty greenish yellow. Now add to the 45 c.c. of 
liquid three to five (burette) drops of the methyl 
orange used as indicator. This turns red, the colour 
being entirely masked by the deep red of the solu- 
tion ; the colour change with alkali is similarly 
masked, as the beetroot pigment also turns yellow. 
Place the liquid to which methyl orange has been 
added into a tall cylinder and suspend a strip of 
filter paper as previously described. The strip is 
gradually stained a fairly uniform purpHsh red. 
When this has reached a height of about 16 or 18 cm. 
remove the strip, let it drain for a few minutes and 
then dip it into N/25 solution of NaOH, removing it 
immediately. The clear yellow of the methyl orange 
turned by alkaU is very plainly visible at the top, 
over a width of 2 or 3 cm., while the rest of the strip 
still remains purple or red. If the strip is imme- 
diately rinsed, dried and kept in the dark, the result 
may be preserved permanently ; failing this, the 
lower portion gradually gets discoloured. 

Another mixture suitable for demonstrating the 
method is made by extracting 5 gm. of turmeric with 
30 to 40 c.c. of hot water, filtering and adding to the 
filtrate about i c.c. of concentrated picric acid. The 
mixture stains the filter paper a fairly uniform 
yellow ; when dipped in dilute caustic soda, the 
lower portion turns brown, while the upper, which 
contains the acid only, remains yellow with sodium 

The method is capable of very wide application. 
It has been developed, and its possibilities demon- 
strated, chiefly by F. Goppelsroeder, whose work 
unfortunately has appeared chiefly in publications 
not generally accessible. .\ number of papers 
covering a very wide field were published in the 
Kolloid-Zeiischriff, and are given below. 



F. Goppelsroeder, Koll.-Zeitschr., V., 52 (salt solu- 
tions), log (foods and beverages), 159 (mineral waters), 
200 (vegetable pigments), 305 (urine) (1909) ; VI., 42 
(urine continued), iii (vital staining of plants), 174 
(vital staining of animals), 213 (vital staining of animals 
continued), 268 (constitution of dyes and vital staining) 

8- a 

Chapter XVIIl. 


A COMPARATIVELY simple and satisfactory instance 
is the adsorption of oxalic acid by charcoal, the acid 
concentration being determined by titration with 
potassium permanganate. 

Dissolve 10-5 gm. of crystallized oxalic acid 
(C2H2O4 . 2H2O) to make 250 c.c. of solution. Then 
place into five conical beakers or Erlenmeyer flasks 
of 100 to 150 c.c. capacity the following : 50 c.c. of 
the original acid solution ; 40 c.c. of solution and 
10 c.c. of water ; 30 c.c. of solution and 20 c.c. of 
water ; 20 c.c. of solution and 30 c.c. of water, and 
10 c.c. of solution and 40 c.c. of water. Label the 
beakers in the same order, 5, 4, 3, 2 and i respec- 
tively, these being the ratios of their original con- 

Place in each beaker i gm. of finely powdered 
charcoal and shake well at intervals, and finally aUow 
the powder to settle overnight. 

We have now to consider the choice of a suitable 
strength for the permanganate solution. The solu- 
tion (i) has a concentration before adsorption one- 
fifth of that of the initial solution, and this wiU pre- 
sumably be considerably reduced. The final con- 
centration after adsorption should, however, still be 
capable of accurate determination, and we shall, 
therefore, do well to choose a permanganate solution 
so dilute that, say, 25 c.c. of it will be required to 
oxidize 5 c.c. of the starting solution. 


The concentration* of the latter (C2H2O4 . 2H2O 
= 126) is M/3, and we shall, therefore, require 
2KMn04/i5 for complete oxidation of one Utre of 
solution, i.e., 21-086 gm. Since, however, we have 
settled that 5 c.c. of permanganate solution should 
equal i c.c. of the original acid solution, we shall 
require one-fifth of. that concentration, i.e., 4-217 gm. 
per htre. It will not be necessary to make up a 
Ktre, but 500 c.c. (containing 2-108 gm.) should be 
made up with freshly distilled water. This leaves an 
ample reserve for repeating any of the experiments 
which, when the isotherm comes to be plotted, 
appear doubtful ; a reserve of oxalic acid is also 
provided for by the figures given above. 

The first operation is, of course, to determine the 
actual ratio of permanganate solution to oxalic acid. 
Titrate 5 c.c. of the stock solution with the perman- 
ganate solution in the usual way, hot in presence of 
sulphuric acid. Suppose the figure found is 26-1 c.c. 
(instead of 25). Then 5 c.c. of the solutions (4), (3), 
(2) and (i) will respectively require 20-88, 15-66, 
10-44 3-^d 5 '22 c.c. of permanganate. 

Five c.c. of each of the solutions is now pipetted 
off, without disturbing the sediment of charcoal, and 
titrated. Begin with (5), remember the original 
titre, 26-1 c.c, bear in mind that the concentration will 
he perceptibly reduced and, therefore, go slowly. If, as 
will quite possibly happen in one case or another, 
the end point is overrun, the determination must, 
of course, be repeated. Assuming that the sample (5) 
after adsorption requires 17-2 c.c. only, acid equiva- 
lent to 26-1 — 17-2 = 8-9 c.c. has disappeared from 
solution by adsorption. In proceeding to the other 
samples, remember that the amounts which have 
been adsorbed will be smaller absolutely but greater 
relatively. As charcoal varies very considerably, 

* In this chapter "M" is used to denote a concentration of 
one mole (gramme-molecule) per litre. 


no definite figures can be given for guidance. The 
results of a series of determinations, made exactly as 
described, are, however, given below. 

Five c.c. of original acid solution requires 26-5 c.c. 
of KMn04. 

(5) (4) (3) (2) (I) 

5 c.c. of solution requires 
c.c. of KMn04 before ad- 
sorption . . . 26-5 21-2 15-9 IO-6 5-3 

5 c.c. of solution requires 
c.c. of KMnOi after ad- 
sorption . . . 17-5 12-5 7-8 3-8 0-8 

Difference, i.e., amount ad- 
sorbed, in c.c. of KMnO^. 9-0 87 8-1 6-8 4-5 

Since all our units are arbitrary, we can write the 
usual adsorption formula in the simple form : 

y = «C" 

where y is the amount adsorbed and C the equili- 
brium concentration. The latter, expressed in cubic 
centimetres of permanganate solution, is given by 
the figures in the second row, while the figures in the 
third row give the y in the same units. We can, 
therefore, plot the C as abscissae and the y as ordinates 
on sectional paper to a convenient scale, say i c.c. = 
I cm. The points so obtained should lie on a smooth 
curve of the famiUar parabolic type (Fig. 19). If 
any points fail to do so, the corresponding deter- 
mination should be immediately and carefully 

Although the curve obtained may be smooth and 
have the general appearance of the adsorption 
isotherm, it is not possible to say definitely that it 
conforms to the equation without a further test. 
If we take the logarithms on both sides, we find : 

log y = i/n log C -f log a, 

which, taking log y and log C as co-ordinates, is the 
equation of a straight line. To test the nature of the 
curve we must, therefore, plot the logarithms of 



y and C as ordinates and abscissae respectively ; this 
can be done by plotting the actual figures' on logarithmi- 
cally ruled paper, or, if this is not available, by taking 
the logarithms and plotting them to a convenient 
scale, say o-i = i cm., on ordinary milHmetre paper. 
The logarithms in that case should be taken to three 

figures, with the last figure corrected. We thus 
obtain the following values for the results found 
above : 

(5) (4) (3) (2) (I) 

log C . 1-243 1-097 0-892 0-580 0-903-1 =-0-097 
log y . 0-954 0-940 0-908 0-833 0-653 

These values have been plotted, log C as abscissae 
and log y as ordinates, in Fig. 20, and He very nearly 
on a straight fine. The deviation is not greater than 
appears in most of the log y — log C curves to be 
found in the hterature. Whether it is due to experi- 
mental error or actually denotes a departure from 



the ideal type of isotherm can only be determined 
by further test, i.e., by repetition of the titrations 
of (2) and (i), and by determining two further points, 
say one intermediate and one below (i). Suitable 
mixtures would be 15 c.c. of the original acid solution 
with 35 c.c. of water, and 5 c.c. of the original acid 
solution with 45 c.c. of water. As the acid in the 
latter will be almost completely removed, it may be 

^09 y 



-1 1 p 1 1 1 1 r I I 1 II 

0.5 ,._ n "0 

log C 

Fig. 20. 

advisable to carry out the titration with 10 or 15 
(instead of 5) c.c, the result being reduced to 5 c.c. 
by calculation. In view of the smoothness of the 
y — C curve, the deviation from the straight line, 
and, therefore, from the ideal isotherm, is probably 

The log y — log C diagram may be used for deter- 
mining the value of n in the equation of the isotherm, 
since n = log C/(log y — log a) is the cotangent of 
the angle made by the straight line with the C-axis. 


Calculated from the straight Une joining the points 
{2), (3), (4) and (5), this in the present case would 
be 5-5, so that i/« = o-i8, which, although low, 
comes well within the range of observed values of 
i/«. Log a is, of course, the value assumed by log y 
when log C = 0, i.e., the length cut off by the straight 
hne on the y-axis ; in the diagram given log a is 
about 0-68. 

Although the adsorption isotherm and the log y — 
log C curve may be plotted with arbitrary units as 
co-ordinates, for any given solute, comparison with 
another substance is only possible if the results are 
expressed as molar concentrations. This, however, 
is simply a matter of calculation. Our original acid 
solution contained M/3 of oxaUc acid. Solution (3), 
e.g., therefore contained M/3 x 3/5 = M/5. After 
adsorption 5 c.c. of (3) required 7-8 c.c. of perman- 
ganate solution. Since 5 c.c. of the original, M/3 
solution, required 26-5 c.c, we obtain the molar 
concentration of (3) after adsorption, x, from the 
proportion : 

M/3 : a; = 26-5 : 7-8 
—X = M/io-2. 

We therefore know that i gm. of the charcoal used, 
placed in 50 c.c. of M/5 solution of oxahc acid, leaves 
an equiUbrium concentration of M/io-2. This enables 
us to compare oxahc acid with, say, another organic 
acid, using, -of course, the same quantities and concen- 
trations, i.e., I gm. of the same charcoal and 50 c.c. 
of M/5 solution of the other acid. 

The example discussed has been chosen as being 
particularly simple for two reasons : the solute can 
be used in fairly high concentrations, and the method 
of titration is a very accurate one. Similar condi- 
tions, if not quite so favourable, obtain with other 
organic acids, the concentrations being determined 
by ordinary acidimetric methods. Adsorption from 
mixtures can be studied when a specific method is 


available for titrating one constituent ; thus adsorp- 
tion from mixtures of oxalic and some other acid can 
be investigated by determining the whole acid content 
acidimetrically, and the concentration of oxaHc acid, 
in a parallel sample, by permanganate. 

In most cases the difficulties are, however, con- 
siderably greater, and resolve themselves chiefly into 
finding analytical methods of sufficient dehcacy to 
determine small differences of small concentrations. 
Thus with many dyes the whole range of concentra- 
tions investigated may be much below o-i per cent., 
while no specific method of titration is available. 
Determinations of this kind have been made by 
colorimetric methods. If the solute is optically 
active, concentrations may be determined by the 
polariscope, provided, of course, that the specific 
rotation does not vary with the concentration, a point 
which must be ascertained by experiment with a few 
solutions of known strength and approximately covering 
the range to he investigated. 

In determining the adsorption curve the assump- 
tion is made that an equilibrium has been reached. 
Although this is, roughly speaking, true in many 
cases, numerous instances are known in which small 
amounts of solute continue to disappear from the 
solution. The effect of this continued sorption may 
show itself even in the time which necessarily elapses 
between the first and the last titration, i.e., the 
values found for the samples last examined are some- 
what higher relatively than those for the first, a 
discrepancy which would show itself particularly in 
the log y — log C curve. If there is reason to suspect 
this phenomenon, the liquid should be left on the 
adsorbent and determinations repeated at intervals 
of some days. The causes may be various, e.g., the 
adsorption on the coarse external surface of the 
adsorbent is followed by slow diffusion into the pores 
with further adsorption on the surface of the latter ; 


or chemical action may follow adsorption, a possi- 
bility which, although apparently remote, has been 
proved real in some instances ; or, finally, the 
physical condition of the adsorbent, and, therefore, 
its specific surface, may change. 

For theoretical work finely powdered charcoal, 
especially blood charcoal, is the most satisfactory 
adsorbent ; fuller's earth, china clay, etc., give 
discrepant results more frequently. Whatever adsor- 
bent is selected, an amply sufficient quantity to 
carry out and repeat experiments should be obtained 
before starting. 


For adsorption by solids from solution, see H. Freund- 
lich, " Kapniarchemie," Leipzig, 1909, pp. 145 — 173. 
Recent papers : G. v. Georgievics, Monatsh. f. Chem., 
34. 733 (1913). 3-ds. of acids by wool ; T. Oryng, KolL- 
Zeitschr., XIV., 14 (1913), negative ads. ; K. Estrup, 
Koll.-Zeitschr., XIV., 8 (1914), ads. of electrolytes; 
Sorption of iodine by carbon, J. W. McBain, Trans. 
Faraday Soc, XIV., Part 3 (1919) ; C. Koch, Koll.- 
Zeitschr., XXII., I (1918), adsorption of sodium auri- 
chloride on charcoal ; a good example of adsorption in 
very low concentrations. 

Chapter XIX. 


R. E. Liesegang's original prescription is as 
follows : 4 gm. of gelatin is dispersed in loo c.c. of 
water in the usual way, and 2 c.c. of a concentrated 
solution of potassium dichromate added to the sol. 
The mixture is poured on clean glass plates to form 
a thin layer, about 0-45 c.c. per square inch of surface 
being allowed. The plate is supported on a hori- 
zontal surface and the sol allowed to set ; 10 to 15 
minutes will be required, according to the temperature 
of the room. A large drop of 20 to 30 per cent, 
solution of silver nitrate is placed in the centre of the 
plate, preferably by allowing five successive drops 
of about o-i c.c. each to fall on the same spot from a 
small pipette drawn into a sufficiently fine point. 
If this operation is properly carried out, the drop 
should have a clean circular outline. The plate is 
kept in the dark for 24 to 48 hours (as light acts on 
gelatin containing dichromate) , but may be examined 
from time to time in diffuse light. At the end of this 
period any traces of the original drop still remaining 
may be removed with a pointed strip of filter paper, 
and the gel is then allowed to dry. The silver 
chromate resulting from the reaction will be found 
to form numerous concentric circles round the edge 
of the original drop, separated by clear zones free 
from precipitate and increasing in width from the 
centre outwards. 

The following details should be noted. The plates 
must be quite clean and, in particular, free from 
traces of grease. To cover them with gelatin right 


up to the edge is an operation requiring considerable 
practice, and the beginner may be satisfied with a 
uniform layer extending to within |" of it. The 
plate should be slightly warmed and held in the left 
hand an inch or two above the horizontal surface on 
which the plate is eventuaUy allowed to cool, while 
the whole amount of sol is poured slowly on the 
centre and uniformly spread by sHghtly inclining the 
plate as may be necessary. After cooHng and before 
putting on the silver nitrate the plate should be placed 
where it can be left undisturbed for the rest of the 
time, as the drop easily spreads if the plate is moved. 

To produce really good rings the gelatin must 
contain a small amount of acid and of gelatose (a 
product of hydrolysis which does not gelatinize on 
cooling). Inferior commercial brands of gelatin 
happen to contain these two constituents in the right 
proportion, while particularly " hard " gelatins may 
require a slight addition of either or both. Liesegang 
recommends citric acid as particularly suitable, and 
the addition of 5 to 10 drops of a 5 per cent, solution 
to 100 c.c. of sol may be tried if a particular brand 
of gelatin does not give good rings. Similar quantities 
of gelatose may also produce marked improvement ; 
it may be prepared by prolonged boiling of a 10 per 
cent, gelatin sol (evaporated water being replaced), 
which is continued until a sample placed on a cold 
glass surface no longer sets to a jelly. Suitable pro- 
portions of either or both constituents increase the 
width of the chromate rings until, with excessive 
amounts, the whole precipitate forms a continuous 
band. Instead of adding gelatose it may also be 
produced in the sol itself by keeping it at high tem- 
perature for several hours ; the dichromate must, 
of course, not be added until this operation is com- 
plete, as it would undergo partial reduction. 

The experiment may also be carried out in a some- 
what different manner. A test tube is fiUed to about 


two-thirds of its height with the dichromate-gelatin 
sol, which is allowed to set, and a few cubic centi- 
metres of the silver nitrate solution is then poured on 
top of the gel. Other reactions, however, give better 
results with this procedure, among which the follow- 
ing are particularly suitable for study : — 

Tricalcium Phosphate in Gelatin. — Dissolve 3 gm, 
of crystallized tribasic sodium phosphate (Na3P04 . 
12H2O) in 100 c.c. of distilled water and pour the 
solution on 10 gm. of gelatin. Allow the latter to 
swell for three to four hours, then disperse on the 
water bath at 100° C. and filter at 80° to 90°. Even 
the best brands of gelatin give a precipitate with the 
phosphate, but the procedure prescribed makes it 
coarser than it would be if the sodium salt were 
added to the sol. Test tubes, \" or f " diameter, are 
filled with the filtered sol to about two-thirds and 
allowed to cool slowly. The sol must be poured 
slowly down the side of the test tube, to avoid the for- 
mation of froth or bubbles. After the tubes have 
stood for at least one hour, any of the following 
solutions may be poured on, all of which give 
numerous excellent stratifications : 10 per cent. 
CaClg ; a mixture of two parts of 10 per cent. CaClg 
and three parts of 10 per cent. NaCl solution ; 20 per 
cent, crystallized calcium nitrate (Ca(N03)2 . 4H2O). 
Formation of strata continues down to the bottom 
of the tube and is complete in seven to ten days. 

Lead Iodide in Agar. — Dissolve 4 gm. of potassium 
iodide in 100 c.c. of i per cent, agar sol, prepared and 
filtered as described above. Pour the sol into test 
tubes exactly as explained in the preceding para- 
graph and allow them to cool slowly. When they 
have reached the room temperature, pour on a 
30 per cent, solution of crystallized lead nitrate 
(Pb(N03)2). The reaction proceeds rather rapidly, 
and the first, very fine, strata will generally be 
visible in the course of one hour. 


Lead Chr ornate in Agar. — For this experiment the 
agar has to be carefully purified in the following 
manner : Place i gm. of shred agar in a weighed 
200-C.C. beaker and soak it in three changes of dis- 
tilled water, allowing eight hours or thereabouts for 
each change. After the last lot has been poured off, 
the total weight of water (a good deal has been 
•iinbibed by the agar) is made up to 100 gm. [i.e., the 
total weight to loi gm. + weight of beaker), the 
agar is dispersed on the boihng water bath and 
o-i gm. of crystallized lead acetate (PbA'2 . 3H2O) 
dissolved in it. The sol is strained through a plug 
of glass wool and filled into test tubes as before. 
After coohng, a solution of 0-5 gm. of potassium 
dichromate in 100 c.c. of water is poured on. The 
stratifications, owing to the low concentrations, are 
very dehcate, but exceedingly numerous and regular. 
They form throughout the length of the test tube and, 
as the dichromate is in large excess, the gel is coloured 
a faint yellow. 

As agar does not adhere to glass, trouble is occa- 
sionally caused by the aqueous solution creeping 
between the glass and the gel. This may be pre- 
vented in the following manner : a sol containing 
10 gm. of gelatin and 3 gm. of potassium dichromate 
in 100 c.c. of water is prepared. The test tubes to 
be used are filled with the sol, emptied with constant 
turning round their axis, so that a uniform coating 
of gelatin is left, and allowed to cool with their open 
ends downwards. They are then exposed to direct 
sunlight or full daylight for several hours, during 
which time the gelatin coating dries and becomes 
insoluble. Finally they are filled with water, which 
is changed until it remains quite colourless, emptied 
and dried. Agar adheres perfectly to the tanned 
gelatin surface obtained in this fashion. 

Many other reactions may be studied in either 
gelatin or agar gels, particulars of which will be 


found in the literature. The following points should 
generally be remembered. If the aqueous solution 
is to diffuse into the gel at all, its molecular concen- 
tration must be in excess (generally considerable) of 
that in the gel. The concentration need not, how- 
ever, be exclusively due to the reacting solute, but 
may be partly made up by an inert salt. Thus in 
the tricalcium phosphate reaction, solutions of cal- 
cium chloride alone, or mixtures of calcium and 
sodium chloride, give good results ; in the latter the 
concentration of CaClg is lower, but the total molar 
concentration, CaClg + NaCl, is as high, or higher, as 
with CaClg alone. This particular expedient always 
deserves trial when solutions at reasonable concen- 
trations do not give good results, or when salts of 
low solubility have to be tried, which do not diffuse 
into the gel with sufficient rapidity at the highest 
attainable concentrations. Differences in the quali- 
ties of the gelatin used, and in the procedure adopted 
in preparing the gels, may affect the results pro- 
foundly ; this is particularly the case when the salt 
dissolved in the gelatin is not neutral in the con- 
centrations employed, e.g., Na3P04. Agar is less 
variable and is also much less affected by many sub- 
stances which attack gelatin, such as acid or alkali 
liberated by hydrolysis, so that it is to be preferred 
when possible. A reaction which gives good results 
in gelatin, however, generally does not do so in agar, 
and vice versa ; thus Liesegang's reaction does not 
lead to good stratifications in agar, while the lead 
iodide reaction does not produce them in gelatin. 

It is nevertheless sometimes possible to obtain 
stratifications with a combination that does not 
produce them directly, by the intermediate formation 
of a reaction product which appears in that form. 
Two examples may be tried as follows : — 

Liesegang's Silver Chloride Rings. — Disperse 2 gm. 
of gelatin in 20 c.c. of water and add i c.c, of 20 per 


cent, solution of AgNOg. Cover a glass plate, about 
5" X 7", with the sol and allow it to set. Then place 
on it, in the manner described for the silver chromate 
experiment, a large drop of 20 per cent, solution of 
NaCl. The latter diffuses into the gel and forms 
AgCl, which is deposited as a continuous zone. If, 
however, a few small grains of silver chromate 
{i.e., the precipitate of varying composition obtained 
by mixing solutions of dichromate and silver nitrate) 
are placed at points about 10 to 15 mm. from the 
edge of the original drop, rings of silver chloride are 
formed beyond them as the NaCl diffuses to that 
distance. This is, of course, due to the formation of 
sodium chromate and dichromate, which diffuses 
into the gel containing AgNOg, with the formation 
of the usual rings, which, however, are transformed 
into AgCl when the solution of NaCl reaches them. 

Lead Chromate in Gelatin. — Begin the Liesegang 
experiment exactly as described. When a ring of 
silver chromate 2 or 3 mm. wide has formed, remove 
the drop of silver nitrate completely with blotting 
paper, without spreading it, and replace it by a drop 
of 30 per cent, lead nitrate solution. The lead 
replaces the silver in the chromate formed, while the 
resulting AgNOg diffuses ahead, forming fresh rings, 
etc. Lead nitrate placed directly on the dichromate 
gelatin does not form rings, hut only a continuous hand. 

Reactions in Silicic Acid Gel. — Two methods are 
possible. A silicic acid sol is prepared in the manner 
described above, and the one reaction component 
dissolved in it to the required concentration ; the 
sol is then fiUed into test tubes and allowed to set. 
This method is attended with several difficulties. 
Some salts, e.g., iodides or thiocyanates, retard the 
setting very considerably. Repeated heating of the 
sol, but not to boihng, may reduce the time required 
for coagulation. Other salts, such as carbonates 
and phosphates, promote setting to such an extent 

L.M. g 


that they frequently cannot be dissolved completely 
before it occurs. 

The alternative method is to decompose the 
sodium silicate with the acid of which it is desired to 
form insoluble salts, and to use directly the gel thus 
obtained, which, of course, contains the sodium salt 
of the acid used. A lo to 15 per cent, solution of 
crystallized sodium silicate is a suitable starting 
material (prepared with hoiled distilled water and 
filtered, if necessary). A dilute acid is then pre- 
pared, containing in a given volume approximately 
the amount necessary to decompose the sodium 
silicate, calculated as Na4Si04, contained in the same 
volume of silicate solution. A preliminary trial is 
then made by adding to a known volume of the 
dilute acid some methyl orange and titrating with 
the sodium silicate solution until the mixture is just 
neutral. It is then set aside and allowed to coagu- 
late ; if coagulation occurs within a reasonable time, 
say 12 hours, the ratio of acid to silicate may be 
adopted. The necessary quantities of dilute acid 
and sodium silicate are thoroughly mixed (of course 
without the addition of any indicator), and the 
mixture poured into test tubes and allowed to set, 
when the aqueous solution is poured on. The 
following give fine results : gel obtained by decom- 
position with HCl and, therefore, containing NaCl, 
on it 25 per cent, solution of Pb(N03)2 J %^^ obtained 
by decomposition with phosphoric acid and con- 
taining sodium phosphates, on it 20 to 30 per cent, 
solutions of CUSO4, BaClg, Sr(N03)2, MnClg, etc. 
Many other combinations will readily suggest them- 

Preservation of Specimens. — The plates obtained 
by Liesegang's method are allowed to dry and may 
then be kept indefinitely in a dry place. If exposed 
to the atmosphere the silver chromate is, however, 
superficially transformed into sulphide, which shows 


the colours of thin fihns. This may be prevented by 
a cover plate cemented on with Canada balsam. 

The specimens in test tubes must be kept from 
drjdng and, in the case of gelatin, from putrefaction. 
It is advisable to harden the latter \vith a 2 per cent, 
solution of formaldehyde. The aqueous solution 
left in the test tube is poured off and replaced by the 
formaldehyde solution, which is allowed to diffuse 
into the gel for three or four days and is then poured 
off. The tubes, as well as those containing agar 
specimens, may then either be drawn out and sealed, 
or closed with corks covered with paraffin or seahng 
wax. If the tubes are sealed off, this must be done 
very slowly so as not to form a vacuum in the 
upper half of the tube, since this causes the forma- 
tion of gas bubbles in the gel, which disfigure the 


Amberger, 39, 40 
Andersen, 88 

Bancroft, 93 
Bechhold, 69 

Chardin, 45 
Coehn, 83 
Coignet, 43 
Cotton, 88 

Estrup, 123 

Faraday, 33 
Freundlich, 93, 123 

Georgievics, 123 
Gericke, 56 

Goppelsroeder, 114, 115 
Graham, 16 
Gutbier, 33 

Hofmeister, 58, 59 

Ishizake, 93 

Jentzsch, 77, 79 

Koch, 123 
Kohlschuetter, 32 

Kruyt, 93 

Lea (Carey), 31 
Liesegang, 50, 124, 125 
Loeb, 56 
Long, 36 

McBain, 123 
Mecklenburg, 80 
Mouton, 88 

Nelson, 43 
Nernst, 83 

Oden, 93 
Oryng, 123 

Ostwald, Wilhelm, 100 
Ostwald, Wolfgang, 29, 36, 72, 

Powis, 36 

Speck, 93 
Svedberg, 35 

Tyndall, 76 

Ubbelohdc, 100 

Zsigmondy, 25, 30, 33, 97 




Agar gels 




liquid interface 

dialysis of 
dried egg 
electrolyte coagulation of . 
heat coagulation of . 
purification of 
Arsenic sulphide sol . 

Brownian movement 

Cadmium sulphide sol . 
Calcium phosphate, strata in gelatin 
Capillary analysis ... 

Cataphoresis .... 

microscopic method . 
U-tube method . 
Collodion ..... 
acetic acid 
ether-alcohol . 


with continuous flow 

of sols showing osmotic pressure 

Electrolytes, precipitation of albumin sols by 

emulsions . 
suspensoid sols 
Emulsions ...... 

pure oil-water 
with soap solution 
Emulsoid sols ..... 

Ferric hydroxide sol . . . 


I ID, 































37. 64, 











































Gelatin .... 

• • • • • 


brands of . 



gels . 

• • • • • 


hardness of 

• • • • 


melting point 

46, 47 

setting point 

47, 48 

sols . 

44. 53 

Gels agar 




silicic acid . 


strains in elastic . 


Gold numbers 


sols by formaldehyde 




various methods . . . . , 


Hysteresis (of setting point) ...... 


Lead chromate (strata in agar) . . . . . 


iodide (strata in agar) ...... 


Liesegang phenomenon ...... 


original formula for . .12 

4, 125 

preservation of specimens 


Lyotropic series ....... 

51. 52 

Mastic suspension ....... 


Melting point (of gelatin) ...... 


Optical methods of examination . . . . . 


Organosols ......... 


Osmotic pressure, sols showing .... 

27, 62 

Palladium sol ....... . 


Paraboloid condenser ....... 


Parchment paper ....... 




thimbles ...... 


tubes ...... 


Phosphates (strata in silicic acid) .... 


Polarizing apparatus for examining gels . 


Protection ........ 


Prussian blue sol . 


Setting point (of gelatin) ..... 


Silicic acid gel ....... 


sol ...... . 


Silver chloride (strata in gelatin) .... 


chromate (strata in gelatin) .... 

, 124 


• 31 

Carey Lea's method .... 

• 31 




Silver sols, Kohlschuetter's method 


by tannin ...... 

• 32 

with wool-fat .... 


Suspensions ....... 


Suspensoid sols ....... 


electrolyte precipitation of 

. 89 

mutual precipitation of . 

. 94 

SwelHng of gelatin ..... 

• 43 

in electrolyte solutions 

• 52 

Thermostat ....... 

. 104 

Toluene regulator ..... 

. 104 

T3Tidall cone ...... 

. 76 

polarization of . 


Ultra-filtration ...... 

. 69 

Ultra-filters for pressure .... 

69, 70 

spontaneous .... 

• 74 

for vacuum .... 


Ultra-condenser ...... 


Ultra-microscopic observation with dark-ground condensers 79 



choice of . 


cleaning of .... . 

. 9, 10 

Viscosity measurements .... 


Viscometers, Ostwald's ..... 


Ubbelohde's .... 

. 105 

Water . . . . . 


distilled ...... 


redistilled ...... 


Wool-fat ....... 

• 39 


I, t 1 1 I 


■ t I - « S I" «"